One of the most misunderstood parts about landscape photography is the correct way to fit your entire scene within a photo’s depth of field. Where do you focus? What aperture should you use? You might think that these questions are easy to answer with a hyperfocal distance chart, where you provide your focal length and aperture, and the chart tells you exactly where to focus. There’s only one hiccup — if you want the sharpest possible results, these charts are spectacularly wrong. For most landscape and architectural photographers, that’s a big deal. This article explains everything about hyperfocal distance charts: what they are, why they fail, and where to focus instead.
1) What Is Hyperfocal Distance?
The technical definition of hyperfocal distance is quite simple: It’s the closest point to your camera that you can focus, while still ending up with an acceptably sharp region at infinity (i.e., your background in most photos).
Why did I put “acceptably sharp” in bold? Because it’s way too ambiguous. I’ll go more into that later, but this is the main reason why hyperfocal distance charts aren’t workable — and didn’t even work in the past, regardless of photographers’ changing standards for sharpness over time.
2) What Are Hyperfocal Distance Charts?
Typical hyperfocal distance charts look like this, although there are different ones for every sensor size:
Essentially, you input your aperture and focal length, and they output the closest point you can focus and still capture an acceptably sharp background. It’s not just charts, either; you’ll also find hyperfocal distance calculators and apps that give you the exact same values, but with some more flexibility on the inputs they allow.
But, since they aren’t accurate anyway, you don’t need to worry about any of this.
3) Why Are Hyperfocal Distance Charts Inaccurate?
Hyperfocal distance charts are wrong because their definition of “acceptably sharp” is sloppy and inflexible.
When the first hyperfocal distance charts were designed, someone decided that an acceptably sharp background contained some blur — enough to notice in a medium-sized print — but, all things considered, not a massive amount. After that point, nearly every other hyperfocal chart followed suit.
To be more specific, most hyperfocal distance charts are calculated to give you exactly 0.03 millimeters of background blur. (That’s the physical size of the blur projected onto your camera sensor.) If you’ve ever heard the term circle of confusion, this is all it’s talking about: the size that an out-of-focus pinpoint of light appears on your camera sensor itself.
So, what’s wrong with this definition? Perhaps your first thought is that this particular value, a 0.03-millimeter circle of confusion, happens to be too large for today’s world of high-resolution cameras, large prints, and 4K monitors. If we simply created hyperfocal distance charts with a more demanding value — maybe 0.015 millimeters, or 0.01 millimeters of blur — we’d be fine. Right?
Nope. Not at all.
That’s because the biggest issue with hyperfocal distance charts isn’t that their circle of confusion is too large. Yes, that is a problem, but there’s an even more important one: These charts recommend the exact same focusing distance for a given aperture and focal length, and it doesn’t change, no matter the landscape.
Say that you’re shooting with a 24mm lens, and you want to use an aperture of f/8 (since it’s the sharpest one on your lens). Logically, your focusing point should change depending upon the scene in front of you — whether there’s a foreground element nearby, or whether you’re at an overlook with everything in the distance. But, according to a hyperfocal distance chart, all you need to do here is focus eight feet away from the camera, and you’re set.
That is very, very inaccurate. Instead, the best method is to change your point of focus depending upon the scene. So, if every element of your image is in the distance, focus at the horizon. Or, if there’s a foreground element nearby, focus closer than eight feet (and use a smaller aperture while you’re at it).
If every one of your photos has an “acceptably sharp” background, that’s all it will have. It won’t have the best possible sharpness. It won’t keep your foreground as sharp as possible. All that it guarantees — and all that a hyperfocal distance chart tells you — is that your background will have exactly 0.03 millimeters of blur for every single photo.
So, drat. It seems like hyperfocal distance is a useless topic that won’t help you take sharper photos at all. Right?
Not necessarily. On one hand, it is true that hyperfocal distance charts aren’t useful; that should be fairly obvious by now. But that doesn’t mean hyperfocal distance in general is a bad concept. In fact, there is still a fantastic way to find the right focusing distance in landscape photography. It is also, I might add, quite a bit easier than pulling out a chart each time you take a photo.
Hyperfocal distance is still useful in situations like this, since I have to find the perfect place to focus in order to capture these flowers and mountains sharp simultaneously.
4) The Optimal Focusing Method
Before going into the proper way to find your focusing distance, let’s examine the definition of hyperfocal distance one more time:
It’s the closest point to your camera that you can focus, while still ending up with an acceptably sharp background.
The holdup so far is that “acceptably sharp” has nothing to do with the scene you’re photographing. Is there a way to change that?
Indeed there is. Instead of defining it as an arbitrary, inflexible circle of confusion — no matter how large or small — I propose that an acceptably sharp background is one that is equally as sharp as the foreground. In other words, the background circle of confusion should be exactly the same as the foreground circle of confusion.
That, and that alone, will give you the sharpest possible results across the entire frame. You no longer have to worry about your foreground being vastly less sharp than the background; just focus closer until the two are equally sharp. And, if your “foreground” (the closest element in your photo) is a distant mountain, all you need to do is focus at infinity, and you’ll achieve a blur much smaller than 0.03 millimeters.
The closest element in this photo is quite far away from my camera. Functionally, it’s at infinity. So, why would I focus 8 feet away (which is what the hyperfocal distance chart recommends for a 24mm photo at f/8) rather than just focusing at infinity? Something isn’t right.
I can see some arguments from people who prefer, in a particular landscape, that either their foreground or background is significantly sharper than the other. That’s fair — but this technique is about maximizing your sharpness from front to back. If that’s your goal, as is the case for most landscape photographers, you’ll want the two to have matching levels of sharpness.
There is only one question left: How do you actually find the point that results in equal foreground and background blur? Is it all just guesswork?
Actually, the optimal method is remarkably simple: Find the closest element in your photo. Estimate how far away it is. Double that distance, and focus there. (That’s the real hyperfocal distance, as defined by equal foreground and background sharpness.)
If the closest element in your photo is one meter away, the hyperfocal distance is two meters away. If the closest element in your photo is ten feet away, the hyperfocal distance is twenty feet away.
This is called the double-the-distance method, and it’s something that should be stuck in the head of almost every landscape photographer. Focus twice as far as your closest object. Done.
In this image, the nearest rocks are roughly four feet away from the plane of my camera sensor. So, to capture the sharpest possible result in both the foreground and background, I just doubled the distance and focused at eight feet.
If you’re worried about estimating distances perfectly, don’t be too concerned. First, this is no different from what you’d do with a normal hyperfocal distance chart — trying to focus at exactly fifteen feet, for example — so it isn’t anything new. And, on top of that, you don’t have to be totally accurate. If you focus at 2.8 meters rather than 3 meters, your photo will still be vastly sharper than if you followed a “proper” hyperfocal distance chart in the first place.
Another great thing about the double-the-distance method is that it doesn’t depend upon your focal length or aperture at all. The proper spot to focus in every single landscape, no matter your settings, is double the distance (again, assuming that you want maximum foreground-background sharpness).
The settings you use are still quite important, of course. If your landscape extends from three feet to infinity, and you’re focused at six feet, you wouldn’t want to use an aperture of f/2. But even if you do use an aperture of f/2, you’ll still maximize the sharpness in the scene; it’s just better to use a more typical landscape aperture of f/11 or so instead.
Actually, that’s another important point. Now that you’ve found the best possible spot to focus, what aperture should you use for the sharpest photo? A smaller aperture will provide as much depth of field as possible, but it also decreases your photo’s sharpness due to diffraction.
That’s also a crucial technique to learn — and, once again, there is an optimal answer — but it is too long to describe in this article. I’ve already covered everything in detail in my article on choosing the sharpest aperture. That’s a great place to start.
So, is that it? You simply focus at double the distance for every photo, and you’re set?
Yes indeed. For landscape photography, this method is a fantastic tool to have in your kit. It’s how I focus for every single landscape I encounter, given that I want the maximum possible depth of field. Don’t worry yourself with hyperfocal distance charts, because their definition of “acceptably sharp” isn’t up to par. Instead, focus twice as far as your closest subject, and you’ll be set.
5) Are Lens Aperture Scales Also Inaccurate?
Some lenses (though fewer nowadays) have built-in scales to tell you how much depth of field you’ll get at a given aperture. They look something like this:
People frequently ask me whether it’s possible to use these depth of field scales to focus properly and use the optimal aperture. Or, like hyperfocal distance charts, are these scales also wrong?
In practice, there are a couple reasons why you’d want to avoid using these lens scales as a guide for the best possible place to focus. First, you have to ensure that the distance markers on your particular lens are accurate in the first place. Not all of them will be calibrated perfectly, and it’s very possible that your lens will misidentify how far away it’s focused. For example, it may say that it’s focused at five feet, but it’s actually focused at seven or eight feet instead.
More than that, though, these focusing scales are also calibrated with a 0.03 millimeter circle of confusion in mind. This means that their depth of field markers are — to say the least — quite generous. I mentioned earlier that 0.03 millimeters of blur is fairly noticeable on medium-sized prints, and that’s still true. If you follow the indicators on lenses like this, your horizon and foreground will each have 0.03 millimeters of blur. That’s not a massive or unforgivable amount, but, very often, you can do better.
So, as a whole, I wouldn’t use these scales to focus properly in a landscape. They’re not quite as bad as hyperfocal distance charts, but the optimal method is, like always, double-the-distance.
Here, the closest objects in my frame are some grasses at the bottom of the image. They were only about a foot away from the plane of my camera sensor. So, I focused two feet away and used a very small aperture of f/16.
6) Conclusion
Hyperfocal distance charts are wrong for two reasons. First, their definition of an “acceptably sharp” background has a 0.03-millimeter circle of confusion, which isn’t particularly sharp. And, even worse, these charts don’t change at all depending upon the landscape in front of you. So, they simply aren’t flexible.
Instead, it’s best to define an “acceptably sharp background” as being “equally sharp as the foreground.” That maximizes definition across the entire frame, from top to bottom, and it lets you do away with the issues of traditional hyperfocal distance charts.
Best of all, finding this point — the correct hyperfocal distance — is simple. You have to locate the closest object in your frame, estimate its distance from your camera sensor’s plane, and then double that distance.
For landscape photographers, this information is essential. If you’ve ever been at a scene with a great foreground and background, but you’ve been unable to capture both as sharp as possible at once, this is the proper hyperfocal distance. Forget charts and calculators; forget depth of field scales on your lenses. By focusing at double the distance, you can maximize the sharpness of a scene without compromise, and it is far easier to put into practice, too.
Here, the nearest object in my photo is the grass at the bottom of the image. It’s about five feet away. So, I focused ten feet away, which lines up roughly with the front of the island in the middle of this stream.
If you have questions about hyperfocal distance charts, depth of field, doubling the distance, or anything else I covered, feel free to ask about it below. This is a higher-level topic, but it’s an important one that every landscape photographer should know. There’s certainly enough misinformation floating around online about hyperfocal distance, so I’ll do my best to address any concerns in the comments section.
Excellent explanation, thanks!
I’ve been stacking a Near and Far image combination, which is only useful when there is nothing moving.
That is a very useful technique indeed! But yes, for many landscapes it simply isn’t possible. Whenever I can, I always try to get everything as sharp as possible in a single photo, which is where the double-the-distance technique comes in handy.
I am probably missing something but for me using the hyperfocal technique was always double-the-distance technique as you call it. Not sure how else you could do hyperfocal…
I believe Spencer is saying you should use the double-the-distance method instead of using charts.
Informative article, thanks for the education on acceptable sharpness.
I feel as if there is one piece of information missing; what aperture to use in a given situation. You demonstrate this in the last and second last photos. Let us examine the very last photo where you indicate that the nearest item of interest is five feet away and therefore you focused ten feet away. I get the impression that you referenced a hyperfocal chart in order to determine that f/9.0 was sufficient to provide the foreground grass and background mountain in focus. Since we are always fighting diffraction at higher apertures, minimizing the necessary aperture is seen as a benefit and you did not use f/16 as demonstrated in the second last photo where the nearest focal point of interest was one foot away.
Andy
Thank you, Andy! Yes, I considered adding that information into this article, but it was already long enough that I didn’t want to overload it with information.
The sharpest aperture for a photo also does have a mathematically accurate answer. You’re correct, too, that it involves looking at a chart. However, rather than a traditional hyperfocal distance chart, you need to consult a sharpest aperture chart, or a sharpest aperture calculator. Not very many people have written about this in the past. The two best options out there are: First, the article I wrote on choosing the sharpest aperture (https://photographylife.com/how-to-choose-the-sharpest-aperture). Second, the trio of apps written by George Duovos (which I link to under the “6. Smartphone Apps” section of the sharpest aperture article). Most other “how to choose the sharpest aperture” articles out there skirt around the issue and recommend just defaulting somewhere from f/8 to f/16.
Great addition, cheers :)
Hey this is great but instead 0.03 I was using 0.02 which was better but I will give this a try, but when you focus do you use manual focus at double the distant?
Hi Chris, although using a circle of confusion of 0.02mm will improve background sharpness, it doesn’t do anything to improve the lack of flexibility of hyperfocal distance charts in the first place. For example, even with the 0.02mm background blur example, your suggested charts would tell me to focus 12 feet away with a 24mm lens at f/8 — even if the landscape is all the way at infinity (like the mountain example photo in this article).
Doubling the distance works no matter how far away your subjects are. If your subject is 1000 feet away, focus 2000 feet away — both of which are already functionally at infinity, so just focus at infinity.
As for manual versus autofocus, it doesn’t really matter. Double the distance is double the distance. In a best-case scenario, autofocus and manual focus will find exactly the same point, so there will be no difference between them. However, if autofocus is having a hard time locking on your subject, or there is no clear subject at double the distance, manual focus will certainly be preferable.
Understood thank you very much;)
Spencer,
Great article. This is incredibly useful for backcountry use where referring to a hyperfocal chart can be a hassle at times.
Question: Why is the distance doubled rather than tripled, quadrupled, etc? Maybe the answer is obvious, but I’m not seeing it yet.
Thanks!
Mac
Mac, That’s a good question and the answer isn’t obvious: see my comment number 23.
I’m of the thought that having everything appear in focus on a landscape is just a bit unnatural. I actually like having a bit of the background blur out as it better matches what I see. Feels right.
Just a personal preference.
That’s a fair approach. Even in that case, though, I caution against using hyperfocal distance charts. It’s probably best to use the double-the-distance method, and then back up your focus ring slightly to favor the foreground. The reason is that, in a landscape with a massive depth of field (objects nearly touching your lens), following hyperfocal distance charts will give you a 0.03mm circle of confusion in the background — but the circle of confusion in the foreground could be far higher, with areas that are wildly blurry. The DTD method will equalize foreground and background sharpness, and then you can modify it in whichever direction suits your preferences.
Wow! This is something I am going to try the next time I am out shooting landscape. “Focus twice as far as your closest object. Done.” I love it! Thank you for sharing this!
Glad you like it! It’s a surprisingly simple technique, which is why I’m surprised it isn’t known so widely.
A presentation like this is rather incomplete IMHO if it doesn’t also include the alternative approach described in detail by Harold Merklinger: http://www.trenholm.org/hmmerk/
In a nutshell, IIRC, focus at infinity using an aperture setting that is the same size of what you want to resolve at infinity. All you are giving up vs the hyperfocal distance approach is a factor of 2 at the near distance.
I hardly ever use hyperfocal and prefer very much to focus on the detail in the landscape that I want to be in focus. I do like the idea presented in the article above to focus on twice the distance of the closest subject that I want to have in focus. Quite convenient.
I am also wondering why there is not a single camera that has the option to tell it to go to the hyperfocal distance directly. The info is in the EXIF data so it should be easy to provide the option to actually have that value used and set on the lens.
Hi Dieter,
If Harold Merklinger’s method works for you, go for it, but I don’t suggest it in general; it really is a spectacularly bad way to focus. It recommends that you focus at infinity for every photo. That could work without too many issues if your subjects are all far away, but it would be a very poor choice for something like the second-to-last image in the article above, where my nearest subject was one foot away. Not to mention the complexity of this method, and I really see no reason why anyone would do it rather than simply doubling the distance…
Well, it did away with the hyperfocal distance issues (which became worse once lenses no longer had DOF scales). But I do agree, nothing is simpler than the easy message you described in your post. I will read your article one more time to make sure I got it right; Merklinger and hyperfocal never really worked for me.
That’s a fair point. I do see where the method originated; it just isn’t optimal if you want the sharpest possible front-to-back image. Thanks for mentioning it, though — I’ve heard a few other photographers bring it up in the past, so it’s clearly a technique that has entered some degree of mainstream use. But there are indeed better options out there.
Best of luck using the DTD method. I think you’ll find it much easier to implement.
This is always a very interesting subject, thank you for a straightforward ‘rule of thumb!
I don’t think that Merklinger “recommends that you focus at infinity for every photo.” His Rule of Thumb 5 states “If we want all objects of, say, 5 millimeters diameter to be recorded—at whatever distance—we should use a
lens aperture of 5 millimeters or smaller and focus no closer than half the distance to the furthest object.” This is in terms of object disk of confusion rather than image circle of confusion; for instance, if your penultimate image was focused at infinity then all objects larger than 1 mm in size will be resolved regardless of distance (up to overall system resolution).
Maybe it’s sound practice to take 4 images: focused at closest object, twice closest object, half distance to furthest object and infinity. Hopefully one will provide the desired balance of focus!
In practice, the Merklinger method does suggest that you focus at infinity for nearly every photo. As you said, “focus no closer than half the distance to the furthest object.” So if the farthest object is a mountain 3000 meters away, it suggests that you focus at 1500 meters or farther — which, functionally, is infinity. Since most landscapes have some distant object on the horizon, it’s fair to say that the Merklinger method nearly always will recommend infinity focus.
It is true that if you want the sharpest image from front to back, focus stacking several images could be the way to go, keeping in mind the standard precautions about focus stacking. But in a single image, you can only capture the maximum front-to-back depth of field by focusing at the point where the foreground and background are equally sharp — the point where their circles of confusion are the same size. That just so happens to be double the distance to your closest object (although there are a few other ways to find it as well).
I noticed the hyperlocal distance in the EXIF just yesterday while checking my shutter count. While the idea of having the camera automatically focus at that distance, the same COC principle applies and that, as far as I am aware, is an unknown value for what my camera uses.
The hyperfocal distance value in your EXIF is calculated, not measured. It’s calculated from your aperture, focal length, and a CoC of 0.03 millimeters. So, it really is no different than simply looking up the value on a “traditional” hyperfocal distance chart yourself. Not as useful as it might seem, unfortunately.
Thanks for confirming the CoC value Spencer; kinda what I figured but it was only an assumption on my part.
Merklinger assumes that infinity is important (if it is not why bother) and having it being the sharpest then use the aperture to resolve the appropriate size of detail on a nearby object works well psychologically for a lot of landscape viewers. Hyperfocal with putting the nearby subject in total focus then change aperture for an acceptable infinity also works well because you are focusing on a true subject. In both end-emphasis approaches, something that can be identified as important – either a nearby object or infinity – is in best focus, and the gradual change of sharpness is monotonic (one-way). This may look more ‘natural’ and less jarring, and certainly is not spectacularly wrong. With double the distance, potentially something that is not identified as a subject is in best focus and blur increases on both sides until it hits max at the two most important topics, the nearby subject and infinity. This can look jarring. Double the distance works best in certain situations. You may not have anything at the best focus distance – this is the ideal case, e.g. looking across a valley with the photo including nearby objects and the other side of the valley with the double distance point corresponding to transparent air only. The other use is demonstrated in this article, where the entire frame from near to far is regarded as of equal importance and a small enough aperture is used so the best focus point is not obvious from looking at the photo at intended display size and viewing distance – everything looks sharp and nothing is sharper than anything else. In these cases double the distance allows the use of the largest possible aperture to achieve a set acceptable out of focus blur across the frame, mathematically. Compared with Merklinger, you might find that the degree by which Merklinger is worse for the close object is pretty small, and the degree by which Merklinger is better for infinity is quite often more noticeable. Test with photos and see.
DTD = KISS
Thank you!
Yes indeed! I don’t know if there are any simpler methods out there.
:) Just because, about 40 years ago an old timer told me that he just focused on his subject and then pulled the focus in “a tad”. I think that is simpler and is somewhat effective though I have to try your idea as it seems it will be more consistent.
So when you focus at double the distance, you just swivel your camera down and focus on a rock, or a plant, or whatever is at that distance, then recompose and shoot, is that it? Talk about KISS!!!
Yes, that’s right! If you like, you don’t need to focus and recompose, either; you could just compose and then focus by changing your autofocus point (generally in live view), or just manual focus.
Awesome!! Thank you. Great article; I’m all about simple. 🙂
The double the distance method was actually first discussed on PL three years ago in the Forum section:
https://photographylife.com/discussions/topic/getting-all-in-focus-in-landscape
Yes, we’ve talked about it a few times before. It’s a very useful technique! I’ve mentioned it offhand in a few articles, and a bit more in depth in my general hyperfocal distance article (https://photographylife.com/hyperfocal-distance-explained). It looks like a few people talked about it even before that in the link you attached.
We aren’t the ones who invented this technique — not by a wide margin. We just use it ourselves in the field, and hope that other people will pick it up as well.
Spencer, this is a very helpful article. I tend to just shoot at apertures based on past experience when things were acceptably sharp. I tend to be taking photos on the go, but this will be handy when I’m taking more time for a particular shot.
I have to say I really enjoy your photos as well. Nicely done and naturally processed. I know it’s part of the magic, but I’d love to know how you generally post process, especially the second to last shot with the grass in the foreground and cloudy mountain in the background as just one example. I’m always looking for ways to tweak my processing. I generally believe I make it more complicated than it needs to be. I’ve been reading PL for a couple of years now and finally decided to comment.
Hi Andy, thank you for the kind comments. Even with the double-the-distance technique, it’s still true that you have to set a good aperture, in which case past experience helps quite a bit (especially if you prefer not to carry charts, such as the ones I discussed with Andy — looks like a different Andy — in comment 2.1).
I try to do all my processing in Lightroom, if possible. That was the case for the second-to-last image as well. There’s no perfect way to post-process an image — no right or wrong steps you can take to get there. For me, the most important thing is to visualize exactly how I want the final photograph to look, and then make all the edits with that final image in mind (rather than going down the list of sliders one by one and testing which direction I like better).
In that photo specifically, I dropped the highlights significantly and boosted the shadows quite a bit (both of which are very typical in such a high-contrast scene). The grasses in the original foreground were quite dark, and parts of the snow were almost blown out. I adjusted the brightness and contrast sliders very slightly, and I added +19 vibrance. I typically try not to adjust clarity or saturation too much, although this of course depends upon the image. Here, I kept both at zero. Same goes for spot healing — I didn’t do any here, and I very rarely do for anything other than dust specks.
Beyond that were some HSL and local adjustments — mainly to make the grasses more colorful, brighten the corn lilies, and do a bit more selective darkening of the hotspots in the sky and snow. I’ve attached a JPEG export of the unedited file (aside from an identical crop) to this comment. I hope it helps give you a good idea of the magnitude of edits that I made. The most noticeable change should be the brighter grass, which I did via the shadows slider and some local adjustments (including a gradient to brighten them).
Thanks for great info I have queistion please
How do you measure the focus distance ?
what about using laser pointer ?
thanks
For that, I typically just estimate. It should be fairly easy to predict distances, especially at a closer range, and you don’t need to be totally exact in order to capture a very sharp result. If you like, and it’s possible without harming your landscape, you could consider measuring it simply by counting paces. If the closest foreground object is four paces in front of you, walk eight paces (total), and find something on the ground where you are. When you walk back to your camera, focus on whatever object you just noted.
I don’t usually do that, though. Estimation nearly always works just fine.
Liked the advise given by you Spencer. Its an easy thumb-rule to follow. This way, one can decide on what one needs to focus on depending on the scene in front of the camera view. An earlier suggestion I had seen was to focus on 1/3rd the distance in to the frame. But then this would mean that some part may still be in slight blur in immediate viscinity. One can not be accurate anyway about focussing at the Hyperfocal distance anyway.
Yes, the 1/3 focus method has some popularity, but it is quite vague, and not something I’ve ever found to be particularly useful.
Excellent information a really good reason to join this site.
Very glad you enjoyed it!
Spencer thank you for the clear explanation.
I must say that i use an other method, more time consuming:
Looking at the scene in liveview at 100% at the chosen aperture ( and use d810 splitview) and deciding what aperture and focusdistance is best to capture what i want. I have noticed that almost every lens has it quirks that are addressed by this method. It also works with TS lenses.
Sadly the focus scale on most lenses has small throws making the method more cumbersome than necessary
Yes, that is also a good point. If your camera happens to have split-screen focusing, it is a very valuable feature, and it’s true that it makes it easier to correct for lens aberrations like field curvature or focus shift. Thanks for adding this.
great article, many thanks !!
but how can we be sure to use the right aperture depending of the scene?
for exemple, on a landscape I may use f5.6 instead of f5.6 wich be enough, how can we know the perfect aperture for maximum of deep of feild depending of the landscape situation?
Check out the discussion I had with Andy in comment 2.1. There is a mathematically accurate way, but it’s beyond the scope of this article, so you can read about it here instead: https://photographylife.com/how-to-choose-the-sharpest-aperture
The alternative is just to go by past experience and generally just use smaller and smaller apertures depending upon the depth of field you need. But if you’re going for total accuracy, the method in the article I just linked is the proper way.
Ok, you are saying a very wrong sentence: “Hyperfocal charts are inaccurate”. That’s not true. Hyperfocal charts are accurate to determine the correct focus distance with a fixed circle of confusion and aperture and focal length. If this isn’t sharp enough for today high definition photography, that’s another matter , but doesn’t mean that focus charts are inaccurate..
That said, a good article as always.
Regards
Glad you liked the article. I think it’s just semantics. I consider hyperfocal distance charts to be inaccurate simply because they don’t give you the sharpest possible image from front to back. However, if your goal is indeed to capture exactly 0.03mm of background blur, then of course they are completely accurate (and the double-the-distance method isn’t). I don’t see it as having anything to do with today’s high-definition photography, because it still applies even with lower-resolution film; these charts do not recommend the optimal spot to focus if you want your foreground and background both to be as sharp as possible.
Very helpful article for many people, Spencer, and clearly explained as usual. I’ve been using the DtD method myself for quite a few years now, mainly for wide-angle landscape photography.
Obviously, whenever I shoot at middle to longer focal lengths, I focus on the main subject as depth-of-field becomes too narrow to keep both fore- and background elements sharp. (just a little comment I wanted to insert for newbies reading along)
Very true. If it isn’t possible to capture everything within your depth of field, or you intentionally want some out-of-focus regions, it certainly is best to focus on your subject (in most cases) than at double the distance. Thanks for the addition.
Beautiful and awesome.
Glad you liked it!
From Wikipedia: The hyperfocal distance is the closest distance at which a lens can be focused while keeping objects at infinity acceptably sharp. When the lens is focused at this distance, all objects at distances from half of the hyperfocal distance out to infinity will be acceptably sharp.
Clear, concise, and precise. Using the hyperfocal distance doesn’t tell you to focus on the closest object, it tell you to focus at the twice the nearest object.
Any photographer should know this as it been said thousands of time, focus at twice the nearest object WITH the appropriate, hyperfocal aperture. This IS the hyperfocal distance method. The hyperfocal method is designed to keep the nearest and farthest object in equal sharpness.
You can easily see this with this DOF graphing tool:
http://www.tawbaware.com/maxlyons/calc.htm
All you’ve done is clarified YOUR incorrect understanding, and probably confused others.
I’m a bit confused… because we aren’t in disagreement.
You said “Using the hyperfocal distance doesn’t tell you to focus on the closest object, it tell you to focus at the twice the nearest object.”
I said “Finding this point — the correct hyperfocal distance — is simple. You have to locate the closest object in your frame, estimate its distance from your camera sensor’s plane, and then double that distance.”
Throughout the entire article, I never once said to focus on your closest object. (If I did, please let me know where, so that I may correct it.) That would be an awful place to focus, since your background would be wildly blurry. I have always and only said to focus at double the distance to your closest subject. That’s why I’ve been calling it the double-the-distance method.
I hope that clarifies things for you. Let me know if I’m misunderstanding your comment, but I believe we’re on exactly the same page here.
USE BLOODY METRES!!! it’s what the WHOLE world uses! basically it’s pointless article for anybody outside US.
I assume that this comment is a joke, or that you didn’t read the article, but if it is serious:
I used meters and millimeters multiple times throughout this article. I also used feet multiple times throughout this article. I even used paces at one point in the comments section. And that’s because double the distance… is double the distance. It doesn’t matter what units you use.
If you’re talking specifically about the hyperfocal distance chart I provided at the beginning, and my later references to it throughout the article, it’s true that it only has imperial rather than metric units. I didn’t bother providing a meters version because hyperfocal distance charts don’t matter in the first place. If you check out my article on choosing the sharpest aperture (https://photographylife.com/how-to-choose-the-sharpest-aperture) you’ll find that I do provide a meters version of the chart, since — very correctly — you pointed out that much of our audience follows the metric system. That’s something I always keep in mind while writing new content.
I don’t mind critiques of my articles. But to think that this article is “basically pointless” for that reason is quite silly, because double the distance is universally double the distance.
m (metres) ≈ 0.3 × f (feet)
I’ll make that easy enough for a child to estimate using mental arithmetic:
metres ≈ feet times 3, divided by 10
Hope that helps.
Thank you, Pete, I always appreciate your comments :)
Thank you Spencer this is a great article. A simple rule of thumb that is easy to remember. Another recommendation I have see is given there is no particular object of focus, then focus about 1/3 of the way through a landscape frame
Glad you liked it! Do you have any clarity on what the 1/3 focus method is supposed to be? Are you supposed to look through the viewfinder, and go 1/3 of the way up from the bottom (with the horizon line being the top, presumably — or is it the top edge of the frame)? Or is it something else? How would you accomplish that in a landscape like the one below, where most of the scene is above the horizon line, and is in the distance, but there still was limited depth of field (due to my telephoto lens)? In this case, if you draw a line that’s 1/3 up to the horizon, and parallel, it intersects with a hill that isn’t even a constant focusing distance from the camera.
I’ve heard a number of people say to focus 1/3 of the way into the scene. That makes me think it is, to a degree, legitimate, in that it actually does produce sharp results. Done correctly, perhaps it gives you exactly the same result as doubling the distance — since that method (and a few others) already find the accurate spot to focus, and there can’t be more than one correct focusing point in a scene.
Unfortunately, any way I attempt to implement 1/3 focus in practice has always turned out poorly, if I even succeeded in applying it to the landscape in the first place. I hope I don’t sound combative or dismissive, either; your comments on Photography Life tend to be quite accurate, and I simply wish to understand the 1/3 focus method, if it is indeed legitimate, so that I can answer questions more accurately when people bring it up in the future. Perhaps it is the type of thing that works best on a flatter landscape with a wider lens, such as the Death Valley salt flats?
Hi again, Spenser … (Disclaimer: I had a conversation with you in the comments following your related article on this same topic: https://photographylife.com/hyperfocal-distance-explained)
I certainly intend to try the DTD method – with my only reservation being that I’m truly lousy at estimating distance.
Responding to your rhetoric question “Do you have any clarity on what the 1/3 focus method is supposed to be?” ;
Initially, I was assuming (wrongly, as it turned out) it meant one should focus at 1/3 of the *distance* into the scene – which didn’t help me at all (‘cos I was back to square-one with my inability to accurately judge distance).
As I thought more about it, and with some practical experimentation, I found that it meant one should look at the scene in the viewfinder and apply 2 imaginary lines between the bottom or “front” of the scene and the “far back” of the scene (eg. top of the mountain) – so that it’s now visually broken down into 3 parts … then focus at an object that’s on the nearest of the 2 imaginary lines (“1/3 into the scene”).
So, in the last example image in the main section of this article, I’d be focusing on the front-to-middle of the island in the stream – similar to what you concluded with the DTD method. And, in the example just above (tree in front of mountain), I’d be focusing on the top of the foreground hill (front, right) – then I’d recompose for the scene.
You make the point (just above, in your pondering on the “1/3 into the scene” method) that the proposed focus point “intersects with a hill that isn’t a constant focusing distance from the camera” – which I can see is correct … but wouldn’t this be exactly the same situation if the distance to the top of the hill was the focus point determined by the DTD method? (Just wondering, for clarity).
Please note; I’m NOT advocating the “1/3 into the scene” approach over the DTD method – – just explaining it as I’ve worked it out (with the benefit, personally, that it doesn’t require me to estimate distance)
Glad that you worked it out! I do now see what the 1/3 distance people are trying to say. I still don’t have a huge degree of familiarity with that method, but check out Pete A’s comment #23 below to see the technical reason why it isn’t optimal.
If I understand your question correctly, no, you don’t have the same ambiguity with the DTD method. In that landscape, for example, the closest object was the bottom-front of the hill at the right, which was about 10 meters away. So, I looked in the scene for an object that was 20 meters away, which turned out to be roughly at the tree itself (though a slight bit closer).
I hope that makes sense. Let me know if I misunderstood something.
Thanks for your response, Spenser … and I’ve read comment #23 as suggested.
Conclusion: I’m gunna give the DTD method a go – As you point out, it’s simply “double the distance” … so, I don’t have to worry about my [in]ability to judge how far away, say, 15 metres is … I just need to be able to judge the distance to twice as far away as the closest object that I wish to be in focus – without regard to how far it is in any particular metric … and I can confidently do that !!
As I am a landscape photographer, I use a wide angle lens and have my camera set to use one of the buttons to focus and focus disabled on the shutter release. So, what I usually do is focus on something I estimate to be twice as far away as the foreground. This is exactly what is being recommended here although I must admit I was not aware before of any particular reason I should do this.
In effect, the advice here is also not to choose the largest aperture according to the hyperfocal distance but to use the smallest aperture before definition falls off. In effect that means, one should get more critical sharpness from near to far.
I usually use the Nikkor 20mm f/1.8 on my full frame Nikon D610 either at f/8 and f/11, and that gives more than enough range from the foreground to infinity. However, I will do some tests at f/16 because if the lens is still sharp, that might work out even better.
With a a much longer focal length, the range will not always be enough unless one uses a very small aperture, in which case a higher ISO or a tripod may be necessary, possibly both.
Yes, that makes sense. Using apertures like f/8 or f/11 — so long as you have a good mental picture of how much depth of field they provide — is a viable way to choose a sharp aperture for this type of photography. I can’t say that it’s exactly what I do, but, with enough experience, it will provide very good results without the headache of using (or memorizing) charts.
I’ve mentioned this previously, but sadly, it yet again needs to be reiterated:
The common ‘rule’ that one-third of the total depth of field (DoF) is in front of the focus point and two-thirds of the DoF is behind the focus point is a widely-propagated myth. This ‘rule’ is true only in the special case when the focus distance is set to one-third of the hyperfocal distance; for all other cases it is hopelessly incorrect.
E.g., when the focus distance is set to the hyperfocal distance, the DoF extends from half of the focus distance to infinity, which is: 0% in front; 100% behind.
The optimum focus distance is the harmonic mean value of the nearest point (np) and the furthest point (fp):
optimum focus distance = (2 × np × fp) ÷ (np + fp)
When the furthest point is infinity, the optimum focus distance is indeed twice the nearest point distance, as stated in Spencer’s article. Furthermore, as Spencer stated in comment 20.1, this is always true, irrespective of the units of distance measurement.
Thank you for setting the record straight, Pete. I had started to doubt my perspective on the 1/3 method — that it is ambiguous and inaccurate — given the sheer number of people who recommended it. I’m glad to see that my advice in the past had not been misleading, and I appreciate the ability to talk about this subject more confidently and accurately in the future.
Good talking points, Spencer.
The double the distance method is what most experienced landscape photogs do, not the 1/3 of the scene stuff, so glad you shared that. That’s how it was explained for many years here and elsewhere as the practical in the field method.
You are right about the acceptably sharp stuff and how fg and bg sharpness may be different even if at the hyperfocal. If people want to get more critical about the exact hyperfocal, they can check out the advanced calculator here: http://www.cambridgeincolour.com/tutorials/hyperfocal-distance.htm
But at that point, you realize that the only way to get everything really sharp, instead of acceptably sharp, is to focus stack.
However, all of this assumes that we want everything to be equally sharp. In addition, I don’t know anyone who would focus at hyperfocal if they were photographing a mountain in the distance as you mentioned. That just seems obviously wrong.
But what really isn’t covered here is that sometimes a subtle fade off in sharpness, either in the foreground or in the background is actually desirable. It can be used as a tool to enhance the sense of depth and help guide the viewer’s eye from slightly softer areas to slightly sharper areas, just like brightness or contrast. Mindfully and intently choosing what to include and how to guide the viewer’s eye speaks to me more as an artist than robotically aiming to have everything from front to back tack sharp.
Thanks, Spencer.
Hi Spencer,
Pls help clear my understanding b/c I ever watched on line photography course. if I listened correctly the instructor suggested DTD method for taking landscape photo by double the distance focus from the foreground if I want the sharpness from foreground to background. It means if the foreground that I want it sharp one feet away from my camera I have to focus three feet away from my camera (two feet from foreground). But when I read your article it is different from my understanding. For example, let’s see your fouth beautiful landscape photo
“Here, the closest objects in my frame are some grasses at the bottom of the image. They were only about a foot away from the plane of my camera sensor. So, I focused two feet away and used a very small aperture of f/16.
1. you used “the closest objects in my frame are some grasses at the bottom of the image.” Does it mean ordinary grasses in front of the bunch of greenest plants (outstanding foreground), right ?
2. If so, it means if I want it sharp from the first grass that show up in the photo (or seeing in viewfinder) to the cloud. I have to focus two feet away “from my camera (or sensor plane), not from the first grass “, right.
Sorry, If my questions seems like silly or obnoxious
Thank you
No worries, there’s no such thing as silly questions if you’re trying to figure out a new technique.
It seems as though the photographer in the course you watched is incorrect, or else I am misunderstanding you. In practice, all you have to do is estimate the distance between the plane of your camera sensor and the nearest object at the front of your photo. Double that distance, and focus there.
So, if you want sharp grass at the very front of your photo, and the grass is six feet away, focus twelve feet away. The exact numbers aren’t important, though; just focus at double the distance.
To answer your questions individually:
1) Yes, I’m referring to the grasses at the very bottom of the frame. They might not be as eye-catching as the large green plant, but they’re closer to my camera, and I still want them to be sharp. They were about one foot away from the plane of my sensor, which is why I focused two feet away.
2) You’re totally correct.
Hope this helps!
Thank you again for your clarification, Spencer. Now I know why my landscape photos had problems about the shrarpness from front to back, I will use this technique correctly next time.
Great, glad I could help!
Spencer, I went out and took a couple of test shots and I’m sold! !
But this is far too simple it shouldn’t work ! :)
LOL, that’s what my workshop attendees tell me all the time – it sounds too simple, too good to be true! But it always works :)
Never did trust the markings on lenses. Most, as you say, are worked on 1/750 the negative area which is fine for 6x4ins prints from 35mm, but not much more. I find 1/3000 works well for reasonably-sized enlargement. That gives roughly 22ft as a 50mm lens’ hyperfocal [H] distance at f/22 and a 24mm at f/4 works out to a whopping 59ft. For the mathematics, that’s H = Xf/s, where X is the chosen area, f the focal length and s the aperture. The rule-of-thumb used to be a third in front of H and two-thirds behind in focus, though do like your comments on double the distance where not everything needs to be critically sharp.
I agree with Alessio. The charts are not ‘wrong’ or ‘inaccurate’. They are related to sth (lets say to ‘the circle of confusion)’ and are accurate in this way.
Apart from that, most bodies provide a depth of field preview button that helps to control your image in particular when using a tripod and live view.
Finally it depends on the size you publish an image. Since most people publish the image in the web, the original film-aera based charts work very well IMO.
A quick comment on the Depth of Field preview feature – always found that feature to be completely useless on every DSLR. When shooting landscapes, you cannot see squat, especially when shooting at sunrise / sunset. The best way is to use live view / EVF to see how sharp areas of an image are and use the double the distance method to get optimum hyperfocal.
As for film-era based charts, I would never recommend them to anyone. Whether one publishes images on the web or makes large prints, it is always desirable to yield maximum sharpness on a full-size image when sharpness is desired…
Pleas check Christopher’s comment 29) below regarding film-era based charts.
What I find interesting is that most people have forgotten that hyperlocal distance scales were originally created with a known final output size of an image. The 35mm film camera scales were often created to make acceptable sharp 8×10 prints and when you know this the scale can’t be used to make a 16×20. It is stated that Hasselblad depth of field scales were created on the 500 series to make 8×8 inch prints and if you wanted to make 16×16 you need to use two stops smaller scales. It is in the manual. So to state they are simply incorrect what you might say is that most people don’t know how to use the hyperfocal distance scales correctly and understanding circle of confusion and final output size relationships will enhance your ability to interpret the scales. I would also add that in painting distance is enhanced with a loss of sharpness, contrast and if in color with a color shift as objects get farther away from the viewpoint. I can always accept slight sharpness loss at infinity but loss in the foreground is counterintuitive to how we see.
Interestingly, though, the output size doesn’t actually make a difference.
If you think about it, this should make logical sense. There is one single focusing distance, and one single aperture, that gets you the sharpest possible result, regardless of how large or small you print it. It certainly is true that you might not notice inaccuracies in smaller prints, just like how you wouldn’t notice in a smaller print if you accidentally focused on a portrait subject’s nose rather than eyes. However, just because you don’t notice doesn’t mean that you’re focused at the optimal spot.
In practice, the ideal spot to focus is the one where your foreground and background are equally sharp. Now, if you personally prioritize foreground sharpness a bit more than background sharpness (which is somewhat common, if you’re trying to add a sense of depth), that’s fine. Once you find the optimal focusing point, just move it back a bit, to match your tastes.
To find the optimal aperture, regardless of output size, consider reading this article, where I go into much more technical detail about how it all works. If you still have questions afterwards, let me know, but hopefully this should clear up any confusion: https://photographylife.com/how-to-choose-the-sharpest-aperture
Spencer,
of course the output size matters, since the output medium has some resoltion. Maybe ‘output size’ is not exact, but ‘output size and resolution’ matters. Think of a print (any size) with one one pixel as some extreme example. Regardless the sharpness of the original image, all prints will show the same degree of sharpness (which is a very bad one in this example). Or as a more realisitic example, for publishing in the web anything that resovles more than the final JPEG (lets say full HD for instance) doesn’t affect the published image. This applies to the DOF sharpness as well as sensor resolution or lens sharpness. Another typical example are printed newpapers (yes, they still exist) that don’t allow anything more than around 200 dpi since the typical newsprint is too rough. So limited by the paper size and the print resolution anything sharper does not add value.
Regards
That is a fair point. The output size/resolution does matter, in the sense that if your intended output is a 200dpi 8×10 print in a newspaper (just as one example), there is a maximum amount of detail that can possibly be shown, and, whether or not your print has more detail than that, you won’t notice it in the print.
I guess the way I’ve been describing it so far is to maximize sharpness of the initial, full-resolution image, regardless of output size, since it can’t hurt. But it is quite true that the difference between “optimal” sharpness and “very good” sharpness isn’t visible in smaller and lower resolution prints, and, if you don’t intend to print or output larger, isn’t something that you need to be concerned with.
bildsee,
Your comment has reminded me that there are two basic approaches to photography: technology-centric versus client-centric / audience-centric.
For my personal photography, when I press the shutter button I’m not certain how the final image will be displayed. However, each time I consider work for a client, the first thing I establish is the viewing conditions of the final images.
Example 1. Producing the artwork for a compact disc audio album. I made the client sign an agreement that they would not use my images for any other purpose, such as the artwork for a 12 inch vinyl version of the album.
Example 2. My images would be displayed to the public, in a large reception area, via a 48 inch HDTV mounted on a wall, well above head height. Using FX format cameras, this meant that I could get away with using any of the following: f/32 (because nobody will notice the diffraction); a circle of confusion much larger than 0.03 mm; high ISO values (because nobody will notice the noise).
In the above examples, the only difficulty was obtaining the shallow depth of field I wanted in a few of the images. I achieved this using fast telephoto lenses at full aperture. The last thing I needed to obtain was maximum sharpness.
Having learnt photography decades ago, from first principles, enabled me to complete a wide variety of very interesting and demanding projects that the majority of photographers would refuse.
In all of my fields of work, my approach has always been client-centric / audience-centric. I’ve never been technology-centric [What can the technology do for me?]; I’ve always been client-centric [What can I do for my clients?]. New technology doesn’t change any of the laws of physics that govern our universe, however, the latest technology quickly becomes old technology. Whereas, learning to master the underlying principles is acquiring a skill that applies to all of the technology, old, new, and future iterations.
When I need serious depth of field, I use the smallest sensor that will do the job. My Nikon P7800 (12MP) gets more use than my Nikon D600 (24MP) for landscapes. Stitching works, when I need large. This, from a guy who used to shoot Toyo Field 4×5’s and Mamiya 645.
Interesting comment, Chris! I remember long ago, when I switched from a P&S camera to an aps-c dlsr, that the smaller dof bothered me a lot when shooting macro’s.
The notion that smaller sensors provide more depth of field (DoF) is another widely-propagated myth. DoF is an object space parameter, not an image space parameter. Depth of focus is the related image space parameter, which does change with sensor size but this change is irrelevent to DoF.
For any given scene, the position of the lens entrance pupil in 3D object space determines the perspective; and the diameter of the lens entrance pupil determines both the DoF and the level of diffraction. Here’s an example of two camera formats which will produce identical 12×8 inch prints in terms of: perspective; subject-to-image magnification ratio and angle of view; DoF; and diffraction…
135 (FX) Format Camera
sensor: 36 x 24 mm
circle of confusion: 0.03 mm
lens: 50 mm set to f/22
Tiny Format Camera
sensor: 3.6 x 2.4 mm
circle of confusion: 0.003 mm
lens: 5 mm set to f/2.2
Subject distance = 2 metres
DoF near: 0.67 m
DoF far: 2.02 m
Diffraction will be noticable and it will be identical in both prints.
The entrance pupil diameters are identical: 50/22 mm = 5/2.2 mm.
To make the prints also identical in terms of scene motion blur, the shutter speed used on both cameras must be the same. Therefore, the photon shot noise and the overall image-space signal-to-noise ratio delivered to the camera sensors will be identical. If there is a visible difference in noise between the prints then it is caused by nothing other than differences in sensor quantum efficiency and/or read noise.
If the myth that smaller sensors provide more depth of field was true, it would break two of the fundamental laws imposed by our universe: 4D space-time and quantum mechanics!
Pete is entirely accurate. It’s possible to match depth of field between a small sensor and a large sensor without any trouble at all. That goes into another long discussion, this time on equivalence, but the bottom line is that landscape photographers are at no disadvantage with a larger sensor due to depth of field.
I think your method will work well at f/8 or smaller apertures when the nearest object is four feet or further.
There are cell phone apps that take sensor size and megapixel count into account when calculating the hyperlocal distance.
Those apps, unfortunately, are no more accurate than hyperfocal distance charts (unless you’re referring to George Duovos’s apps): http://www.georgedouvos.com/douvos/OptimumCS-Pro_Optical_Science.html
There is only one best spot to focus in any landscape, assuming that you want your foreground and background to be equally sharp, and you aren’t limited by what aperture you can use (i.e., for nighttime photography). That distance is twice as far as your closest object.
There’s also only one best aperture for any landscape, which balances the effects of diffraction and depth of field. That aperture can be found in the apps I just mentioned, or at this link: https://photographylife.com/how-to-choose-the-sharpest-aperture
Hope this makes sense!
Hmmm
am I missing something or am I the only one not understanding how can you focus at a certain distance if not even your lenses scale are to be trusted? Should I look into the liveview / viewfinder and guesstimate something which is at double the distance of the closest object to be sharp? What about the large focusing rings of the Zeiss lenses? are they inaccurate too ?
Your middle idea is correct: go through the frame to select which is the closest foreground element you want in focus, then look for a point that lies about twice as far in the distance, and focus on that one. How you focus on that twice-the-closest-distance element is entirely up to you – if you trust your camera’s AF, then by all means use the AF; if you prefer manual focus then use that. The (in)accuracy of focusing rings (??) doesn’t play any role.
Personally, I just estimate it. I compose my photo, look at the closest object that I included, and estimate how far away it is from the plane of my camera sensor. I then double that distance, searching for something in the scene that is twice as far away. And then, as Greg said, I just focus on that object, either manually or automatically.
“… the biggest issue with hyperfocal distance charts isn’t that their circle of confusion is too large. Yes, that is a problem, but there’s an even more important one: These charts recommend the exact same focusing distance for a given aperture and focal length, and it doesn’t change, no matter the landscape.”
It seems you misunderstood the purpose of these charts, which is to find the hyperfocal distance, which does not change no matter the landscape. The charts SHOULD give the exact same focusing distance because that is what they are for.
Hi Nick,
How do you define hyperfocal distance?
I use the definition: “The hyperfocal distance is the closest point to your camera that you can focus, while still ending up with an acceptably sharp background.”
Next question: How do you define “acceptably sharp?”
Hyperfocal distance charts define it as a 0.03mm circle of confusion in the background. However, I (and many other landscape photographers) define it as equal in sharpness to the foreground, so that the entire image is, from front to back, as sharp as possible.
Hyperfocal distance charts aren’t wrong if your definition of “acceptably sharp” is a 0.03mm circle of confusion. Hyperfocal distance charts are totally wrong if your definition of “acceptably sharp” is to be equally as sharp as the foreground, so that their circles of confusion are identical. And, since that’s what most landscape photographers want, I see nothing wrong with calling hyperfocal distance charts inaccurate for most purposes.
That’s all this article was about. It’s mainly intended for intermediate photographers who hadn’t yet heard of the double-the-distance method, which is just one of many to find the “true” hyperfocal distance — the one where the background and foreground are equally sharp.
The hyperfocal distance is the closest distance you can focus while keeping infinity acceptable sharp, and acceptably sharp means “appears to be in focus” which varies depending on the format and viewing conditions that determine the COC. If you come up with your own definitions for “hyperfocal distance” and “acceptably sharp”, then the charts will not apply as they are not based on your own personal definitions.
My definition for hyperfocal distance is the same one as yours, and Wikipedia’s, and everyone’s who knows what they’re doing.
“Acceptably sharp,” on the other hand, is inherently subjective. But one common definition today is that “acceptably sharp” means that your background is as sharp as your foreground — where both have circles of confusion of equal sizes.
That’s not my own definition. I wish I had invented this idea, but I did not. It’s been known for decades. To get the sharpest photo from front to back, you want your foreground and background to be equal in sharpness. That should seem logical, but it’s not something hyperfocal distance charts tell you.
Instead, those charts tell you how to get exactly 0.03mm of background blur every single time, without any regard whatsoever for what your foreground looks like. They don’t care if your foreground is touching your lens, or if it’s a million miles in the distance. Given the same input focal length and aperture, they’ll always tell you to focus in exactly the same spot. And you’re right; if that’s what you’re trying to find — the focusing distance where the background has exactly a 0.03mm circle of confusion (or substitute in your own value if that isn’t “acceptably sharp” for your uses) — then they indeed serve their purpose perfectly.
But if your goal is to capture the sharpest photo from front to back, they aren’t accurate. I think you know this as well. Instead, the only truly accurate spot to focus is twice as far as your closest subject, which is where your foreground and background will be equally sharp. As for your aperture, the best option is to use a technique I’ve already discussed before (and, again, I’m far from the first one to say this): https://photographylife.com/how-to-choose-the-sharpest-aperture
The important thing to remember is that none of this depends upon print size or viewing distance. Not at all. Because, as should be logical, the sharpest photo is always the sharpest photo. Name a single situation where the sharpest aperture/focusing distance for a 24×36 print is not, also, the sharpest way to capture an 8×12 print. Although you can potentially get away with suboptimal settings on a smaller print without noticing an issue, it should be clear that there is no such situation where the ideal settings are any different.
Hyperfocal distance charts are, in a sense, still useful. They are useful for handheld photography where you’re prioritizing a sharp background (0.03mm or otherwise) over a sharp foreground, and you’re worried about having to raise your ISO in order to accommodate smaller and smaller apertures. In that case, it does help to know your print size, since it will determine how wide of an aperture you can use without noticing background blur.
But these charts were never intended for tripod-based work, where you generally have all the flexibility in the world to set the proper hyperfocal distance and aperture and capture the sharpest possible photo. In cases like that, you can ignore charts that give you a fixed background blur, and instead use the hyperfocal distance that renders your foreground and background equally sharp, with the aperture that best balances depth of field and diffraction.
I hope this makes sense. Thanks for giving me the opportunity to distinguish between handheld photography and critical, tripod-based work, since it’s true that the focusing methods and apertures for each will be different. Let me know if there is anything that you’re still wondering about.
My way is just focus at infinity and accepted the fact that things near the lens won’t be sharp
not a problem for landscape in general
anything wrong with what I do?
Want to hear from you guys
thanks
If you don’t mind having a blurrier foreground, that’s a workable method. However, I definitely recommend testing out the doubling-the-distance method (or any other method to find the same focusing distance). You’ll find that you sacrifice almost invisible amounts of background sharpness, while improving sharpness in your foreground significantly.
Or you could use a smartphone camera , where everything is in focus :-o
Why on earth would people refer to such useless hyperfocal distance charts? The issues brought up in this article are right on when looking at such basic charts, but there are free charts available online that show you the hyperfocal distance relative to the subject distance, and that also allow you to select a much smaller CoC, and show you the near/far in-focus points for any given aperture. At the very least, you could use the good charts along with the double the distance rule to set up correctly.
For a proper DOF chart, take a look at http://www.dofmaster.com/doftable.html
True, in that you shouldn’t look at hyperfocal distance charts, because they are quite inflexible! However, one issue with calculators where you can choose your own circle of confusion is that they, too, are not necessarily optimal.
If you think about it, this should make sense: There is a spot to focus, and an aperture to use, that will give you the sharpest possible photo, no matter what. Output size doesn’t matter.
Hope that makes sense. Thanks for sharing your perspective.
Hello, Spencer !
Excellent explanation….. What about micro 4/3? ” Focus twice as far as your closest object. Done.” How does it work for micro 4/3 ?
thank you in advance for your explanation
Joseph
This method works exactly the same, regardless of your camera sensor size! So, yes, even with micro 4/3 (or a medium format camera, or a one-inch sensor, or a smartphone), you’ll maximize your front-to-back sharpness by focusing at double the distance.
We have an advantage with m4/3 and other cropped sensor cameras of greater depth of field. So I can get approximately the same DOF at f/5.6 that a full frame gets at f/8 for an equivalent lens. This means that I can increase the shutter speed on my Olympus m4/3 or lower the ISO for a hand-held shot. Maybe not a big factor with wide-angle, but more so with tele shots.
That’s a very common misconception, so I’m glad you brought it up. In practice, micro four thirds cameras may appear to have 2 stops more of depth of field (so that f/4 looks like f/8 on a full frame camera). However, they also perform 2 stops worse in ISO performance, assuming identical sensor technologies. So, the effect you mentioned isn’t actually a benefit. At ISO 100, f/4, 1/100 second on a M43 camera, you’ll see a very similar result to ISO 400, f/8, and 1/100 second on a full frame camera, with the only differences being down to the individual sensor differences (such as aspect ratio) and lens performance (such as amount of vignetting). For a more technical explanation of why smaller sensors don’t have a depth of field advantage, check out Pete A’s comment #30.2
Hello Spencer,
I use Hyper Focal Pro mobile app , a free download app for my Android phone ,It asks three parameters after selecting my camera model that is focal length, aperture( f number) and subject distance distance and it outputs two things one is Hyperfocal distance and second is where to focus ( the focus distance)
let me give an example:
Input:
Camera :Nikon D800
focal length: 24mm
Subject distance: 150feet
output:
hfd : 8.248 feet
focus distance : 150 feet
I think it is doing its work very smartly saying to focus at 150ft not to hfd, also says dof is infinity.,
Next I simply changed the input Subject distance to 15feet rest of inputs same, now it shows the same hfd and says to focus to 15feet .
But dof chart given in the article the input is only focal length and aperture not three inputs as in the app, so do you think the app is better than the chart , I am little bit confused and it also shows the cof as 0.028788965mm when I select Nikon D800 camera is that means it is using that value as input to calculate hfd ?.
Your feed back on this is much appreciated and Thanks in advance.
KVS Setty
No, the issue with apps and calculators is exactly the same as with charts. Their definition of “acceptably sharp” nearly always is suboptimal. The ideal answer is “equal sharpness between the foreground and background.”
Instead, simplify things. Focus twice as far as the closest object in your scene. Use an aperture of f/8 to f/16, depending upon how much depth of field you need. Done.
(If you want a more precise recommendation on aperture, I recommend reading this article, but not everyone will want to carry along a bunch of charts: https://photographylife.com/how-to-choose-the-sharpest-aperture)
With a wide lens one can not miss using at least F8 and focusing on the subject.
Using a long lens in most cases a blurred background is far better.
That is a good quick-and-dirty method if you don’t have time for optimal focus. If you have a wide angle lens, and you focus on your foreground subject, chances are good that you’ll get solid (not ideal) results just by using an aperture of f/8 to f/16.
Focusing at the optimal spot (as discussed in this article) can take a bit more time. Still, if you’re willing to put in the effort, and you want equal fidelity in your foreground and background, you can’t beat the sharpness it provides.
Great information. I can’t wait to try it. Thanks.
Best of luck! I think you’ll find it quite useful.
The method I was taught 30-years ago by a professional photographer and artist (FRPS) regarding hyper-focal distances was as follows. Firstly it’s important to understand that the HFD definition describes a distance and f-stop where everything after that distance is in (acceptable) focus. You shouldn’t focus at the HFD because that would put the subject on the border of acceptable sharpness. Instead, you would focus at between half and 2/3 the distance to the HFD. Secondly it’s important to understand that HFD was to be considered only at small apertures (f11/16/22) because not only is the DoF larger there, but also because the lenses in those days performed best at F8-16 regarding both aberrations and sharpness. No landscape photographer would consider f2.8/5.6, and would always be using a tripod so shutter speeds were not an issue. By focussing around half to two/thirds the distance to the HFD, when using a stopped down lens, meant somewhere between 1ft and 35ft away following the chart above (/2) would be in prime focus, where typically the important foreground object would be.
These old ‘rules’ need updating for modern sensors and lenses certainly, but I think it’s still good to learn them properly as well.
It seems to me that the photographer you talked with overcomplicated things a bit!
I certainly agree that learning old techniques is a good way to exercise the brain. However, if you want optimal focus, the method in this article (just focusing at double the distance to your closest object) should provide better results, and hopefully much more quickly. You still need to pick a good aperture, of course, but something like f/8 to f/16 will tend to work well for many scenes (and a more precise recommendation can be found here: https://photographylife.com/how-to-choose-the-sharpest-aperture)
Thanks for adding your perspective.
Thanks Spencer,
Actually the HFD is quite simple if understood correctly. HFD was never intended to be a distance you focus on, although many internet based ‘experts’ have suggested that. It is simply the effective optical ‘infinity’ to consider, normally with wide-angle lenses stopped down. Just like you would hardly ever focus at actual infinity (typically slightly less), you would never focus at the HFD. Instead, knowing what your effective infinity is (e.g. 38′ for 50mm DX at f11), allows you to focus anywhere between that down to half the HFD whilst retaining a DoF to infinity (e.g. down to 19′ for 50mm DX at f11). This is shown on the DoFMaster tables others have kindly linked.
What you can also see from the DoFMaster tables is that the larger the format, the shorter the HFD, which is the other benefit MF and LF sensors still provide for landscape work. At very short HFD (say <3ft), you can focus down to about 2/3 the HFD (rather than 1/2) whilst retaining infinity in the DoF.
This article doesn’t make sense, as a couple of other respondents have pointed out – it is based on a misunderstanding of HFD. “When the first hyperfocal distance charts were designed, someone decided that an acceptably sharp background contained some blur.” No they didn’t, where did this idea come from?
HFD charts calculate the same standard of sharpness in front and behind the subject, extending to infinity – just like the double-distance technique described. The result is exactly the same. Check any HFD chart and you’ll see that the closest point of sharp focus is half the HFD.
The double-distance technique is a good one though, easier and more accurate, because it doesn’t rely on the lens’ distance scale. They’re a) often inaccurate, b) with too few sparsely spaced markings for precise working, and c) they’re measured from the sensor whereas all DoF and HFD calculations are taken from the front of the lens.
Hi Richard, thanks for adding this. I definitely see what you’re saying, and your post really made me think. In a sense, you’re completely correct — “traditional” hyperfocal distance charts will equalize sharpness between the background and half the distance where you’re focused. So, how is that inaccurate? After all, it’s exactly what I’ve been saying about the double-the-distance method. The issue here is quite subtle — more than in any other comment so far. But you are still making a couple misassumptions, and I’ll do my best to clear them up. Bear with me, though, because this one is more complex than most.
Most importantly, just because hyperfocal distance charts equalize sharpness between infinity and half your focusing point does not mean that your focusing point is the same as what doubling-the-distance recommends. In fact, any time that you focus anywhere, infinity and half your focusing distance will be equally sharp. It doesn’t matter whether you used a hyperfocal distance chart, or whether everything in your photo is wildly out of focus in the first place. Even if you have an 85mm f/1.4 portrait lens focused at its minimum focusing distance, it’s still totally true that infinity and half your focusing point will be equal in sharpness. Strange concept, but true.
What this means is simple: You want your closest subject to line up with half your focusing point. That way, it will be exactly as sharp as the background. In other words, focus twice as far as your nearest subject. I think we’re on the same page in that regard.
Hyperfocal distance charts, though, simply don’t do this. They don’t accommodate for how close or far away your nearest subject is. Yes, the region halfway to their recommended hyperfocal distance is equal in sharpness to the region at infinity, but these charts offer no guarantee that your subject will happen to be exactly at that spot.
That is, they offer no guarantee unless you read hyperfocal distance charts somewhat differently than they’re typically used — unless you read them by finding the distance to your nearest subject, doubling that distance, locating that new distance within the hyperfocal distance chart, plugging in your focal length, and only then finding the aperture your chart recommends after all of that. (So, as opposed to plugging in your aperture and focal length to get a focusing distance, you find the right focusing distance via DTD, and then plug in your focal length to get the recommended aperture).
So, let’s work with the assumption that that’s what you’re doing. What would be wrong with that method? Ignoring the fact that far-in-the-distance landscapes won’t even have a hyperfocal distance value on the chart (i.e., landscapes where everything is at infinity), the specific issue is that the recommended aperture is one that results in a 0.03mm circle of confusion in both your foreground and background. Most of the time, you can do better.
How much better? It varies by the landscape. That’s why we can’t just adjust hyperfocal distance charts for modern standards by shrinking their value to something like 0.02 or 0.015 millimeters. The reason it varies by the landscape is that the effects of diffraction come into play, and, if your subject is especially close to your camera, finding the optimal aperture to balance depth of field and diffraction is a tricky thing to do. F/8 will be optimal in some cases, while f/16 will be optimal in others.
Luckily, there’s a mathematical way to find the sharpest possible aperture in cases like this. I’ve linked to this a few times already, but it’s more relevant here than anywhere else: https://photographylife.com/how-to-choose-the-sharpest-aperture
This is an article I’ve already written. It’s an article where the hyperfocal distance charts are a bit stranger than usual; rather than it providing the hyperfocal distance (twice as far as your closest subject) you provide the hyperfocal distance. You also provide the focal length, and it spits out the optimal aperture, balancing depth of field with diffraction. Sound familiar?
I give you full props for your comment. It made me consider this from a totally different angle, and the solution happens to be the exact same one that I arrived at more than a year ago when I wrote my “sharpest aperture” article (and through a completely different path). I really hope that all this makes some sense; please let me know if it doesn’t.
The headline of this article is that hyperfocal distance charts are wrong, and that the double-distance focusing technique puts this right. But they are not wrong, and the double-distance method is simply a good way of setting HFD quickly and accurately – the result is the same.
To be clear, for the benefit of others here’s how it works. Decide on the closest object you want sharp, double that distance by eye and focus on something at that distance. This is your HFD set. Now estimate that distance, refer to a HFD chart (you can’t work without one), and set the aperture that corresponds to it. Job done, and everything from half the HFD to infinity will be sharp.
In practise, HFD focusing only works with wide-angle lenses. While the theory still holds good for longer focal lengths, unless you have apertures like f/64 available, you can’t use them even if you wanted to. The convenient side effect of this though is that the distances we need to estimate and focus on are just a few feet away and easy to judge with sufficient accuracy.
The problem with HFD focusing, and all depth of field calculations for that matter, is not that they’re inaccurate, but with digital working it is easy to break the rules without knowing it and come to the conclusion that they’re wrong – just hit the 100% enlargement button and the whole concept goes out of the window.
The fundamental fact that’s often often overlooked is the whole thing is based on one primary fact which underpins everything. That is, with a 10x8in print viewed at a distance equal to the diagonal (ie 12in) the human eye can clearly discern detail down to 0.2mm, and everything up to that size will appear perfectly sharp. That is the primary circle of confusion and everything is worked back from there.
Fortunately, with larger size prints (or on screen images) the formula is self-correcting and everything will appear sharp provided the viewing distance (equal to the diagonal length) is maintained. And this is traditionally how we view images – small prints we look at quite closely, large exhibition size prints we stand back to view. So far so good, and since our eyes haven’t changed with the move to digital, everything is still valid. Where it all goes wrong is when we press that 100% button and are then looking at an image several feet wide but without changing the viewing distance. Equally, if you view a large exhibition print up close, then the rules have been broken and it will not appear sharp.
Image size and viewing distance are as important when setting HFD and DoF as the lens aperture or anything else. Quick rule of thumb is if the image size is doubled to say a 20in wide print but viewed up close, then the f/number should be two stops higher to compensate.
HFD and depth of field is an optical illusion, and the trick only works if you stick to the rules. DoF only exists as a measurable entity when the image is rendered (as a print, or on screen) and viewed from a certain distance. DoF is NOT set at the moment of capture with an image on the sensor, and this explains why DoF changes with sensor format (and also when you crop an image). There is no way around this, regardless of where you focus – all you can do is move things around a bit, but that’s just a different compromise with other downsides. The only alternatives are a tilt & shift lens that can produce much greater ‘effective’ DoF at lower f/numbers, or focus-stacking with different images blended in software. There are cameras now available that can automate the stacking process, and that has useful potential.
Neither object space nor image space can possess the attribute “depth of field” (DoF), by definition!
DoF is a subjective parameter, it is not an objective parameter. But much more importantly, using 21st-century fundamental logic: DoF is an epistemic parameter, it is not an ontic parameter.
However, Spencer’s article above, and his article to which he linked How to Choose the Sharpest Aperture are ontologically accurate and ontologically precise: his articles are underpinned by robustly-confirmed mathematical and scientific theories.
The extensively-detailed, precise, technical definition of “depth of field” includes a plethora of preconditions and caveats in order to for it to sufficiently bridge the gap between the epistemic and the ontic domains of reality.
Spencer Cox is one of the very few people in human history who have managed to eloquently and succinctly bridge the gap between the epistemic and the ontic domains of reality, in a manner which is easy to grasp and easy to remember for its wide range of intended purposes.
I stated the mathematics (confirmed by the scientific method) in comment number 23:
The optimum focus distance is the harmonic mean value of the nearest point (np) and the furthest point (fp):
optimum focus distance = (2 × np × fp) ÷ (np + fp)
When the furthest point is infinity, the optimum focus distance is indeed twice the nearest point distance, as stated in Spencer’s article. Furthermore, as Spencer stated in comment 20.1, this is always true, irrespective of the units of distance measurement.
Many people suffer from illusions. Very few of their illusions are optical illusions: by far their most frequent illusions are the illusion of explantory power and the illusion of explantory depth.
I’ve fallen foul of those illusions! I intended to write:
Many people suffer from illusions. Very few of their illusions are optical illusions, by far their most frequent illusions are the illusion of explanatory power and the illusion of explanatory depth.
Am I allowed to blame my smell-checker for failing to spot the errors in my worms?
Hi Richard,
You are remarkably close to realizing the same thing I did when I wrote the “sharpest aperture” article, but there is still one thing that appears to be causing some confusion.
Let’s take your quote: “Now estimate that distance [the double-the-distance hyperfocal], refer to a HFD chart (you can’t work without one), and set the aperture that corresponds to it. Job done, and everything from half the HFD to infinity will be sharp.”
Look at this hyperfocal distance chart, for example. It’s calculated for a 0.03mm circle of confusion, as I mentioned earlier:
Let’s assume, for example, that you’re using a 24mm lens, and your closest subject is four feet away. As you just said, that means your hyperfocal distance would be eight feet. So, if you look on the chart, you’ll see that the recommended aperture is f/8.
I will contend, instead, that the sharpest aperture for such a situation is f/13.4 (which, in practice, is f/13). While it does have a greater amount of diffraction, it also has more depth of field, and it results in a background (and equal foreground) circle of confusion of roughly 0.026mm instead of 0.03mm.
That might not seem like a drastic change. But the fact is that the difference here corresponds to an overall sharpness increase in both the foreground and background by 13.3%. People would pay thousands of dollars to get a lens that is 13% sharper than another, and, without a doubt, that difference will be visible in a print.
Where did I come up with these numbers? I simply used the formulae in the article I already linked to: https://photographylife.com/how-to-choose-the-sharpest-aperture. You should also check out this link for similar information and a more in-depth background to the optical science: http://www.georgedouvos.com/douvos/OptimumCS-Pro_Optical_Science.html. And if you go out into the field and test this scenario, I guarantee that you’ll find the same result; f/13 will give you a sharper image from front to back than f/8.
Hopefully, that makes it clear why hyperfocal distance charts can’t be trusted to give you the sharpest result, even if you already focus at the proper hyperfocal distance (the DTD method), and you work them in reverse.
Also, if you’re thinking that you could get an even lower circle of confusion simply by using a traditional hyperfocal chart made with a stricter value, that also is untrue. If you attempt to create a chart with, for example, a 0.02mm circle of confusion, it’s true that your background will be sharper. But there’s not a single aperture out there that will render the foreground of equal sharpness; in this scenario, the four-foot mark will inevitably have a much larger circle of confusion than 0.02mm. The minimum circle of confusion you can possibly achieve in this particular case, with a four-foot nearest subject and a 24mm lens, is 0.026mm (and other landscapes will be vastly different in their smallest achievable circle of confusion).
I’ll end on a final thought: Although it may seem strange at first, the simple fact is that the sharpest aperture and best focusing distance don’t depend upon output size at all. Can you name a single situation where the sharpest settings for a 24×36 print would be different from the sharpest settings for an 8×12 print? Although it will be harder to see optical issues in smaller prints, that doesn’t mean they don’t exist.
I hope all this makes sense.
Make sense? Not really. TBH you’re attempting to reinvent the wheel, that has been turning very smoothly for generations of expert photographers. DoF and HFD are what they are and they work absolutely accurately (within strictly defined parameters, some of which you’ve failed to include). If you accept the concept, and from the foregoing you obviously do, then the science simply isn’t up for debate. Your contention that there’s a difference between the double-distance technique and using a HFD chart is just an error.
To repeat, the DoF and HFD concepts are based on the fact that an average person can discern clear detail down to 0.2mm in a 10x8in print viewed at a distance equal to the diagonal, ie 12in. This is where it all begins, and if you go meddling with the formula what you’re actually doing is changing that fundamental fact from which everything else follows. Do that at your peril, or at least only after exhaustive research that proves it substantially wrong, but what you are suggesting above is that your eyesight is better than average and can perceive finer detail in that print size and distance, down from 0.2mm to 0.175mm, and so we should all accept this new figure. I’m not sure if that was your intention, but that’s what you’ve done. You’ve not covered this crucial point.
There is another very good reason why these things should be left alone, because apart from working very well, they are the basis of a universal standard that has been used for decades in every DoF chart, on every lens scale, and every HFD table. It gives us all a solid platform for comparing apples with apples.
Re your other article on the ‘sharpest aperture’ I don’t accept that either. More reinvention I’m afraid, and also flawed. Diffraction and DoF/HFD are separate things, and while they cross over at higher f/numbers, their impact is different. Raising the f/number increases DoF by extending sharpness at the extreme edges of background and foreground focus. Diffraction reduces sharpness over the whole image, so raising the f/number has the joint effect of improved sharpness at the margins of DoF, but reduced sharpness everywhere else, so cannot deliver the sharpest overall result. Sometimes it might be a good compromise, sometimes not.
Diffraction is one of those things that photographers need to work around according to situation. DoF and HFD settings also have their strengths and weaknesses, and it is common practise for experienced photographers to moderate those too, according to need. By using a higher f/number than suggested for critical working, or by shifting the focus slightly back or forth to ensure that the most important subject areas are as crisply detailed as reasonably possible. It would be great if there was a one-size-fits-all solution that we could just plug in, but there isn’t.
I’m sorry that this doesn’t make sense, but I don’t think there is much else I can do to provide you with extra information at this point. I hope that you’ll read the links I sent you, and that eventually you’ll realize that there is, indeed, exactly what you say would be great: a universal solution to the problem of finding the sharpest aperture for a given landscape. It is a problem that has been solved since at least 1996, when Paul Hansma wrote an article in Photo Techniques magazine on how to determine the optimal aperture given a particular focusing distance and focal length. I suggest that you also read this link, although it is a bit more difficult to parse through formatting and terminology than the other two articles I linked. This is the original source of this information: http://www.largeformatphotography.info/fstop.html (In particular, go down to the “Finding the optimal f-stop to balance defocus and diffraction” header.)
To your paragraph about my eyesight: I don’t quite see where you’re coming from, or where you found the 0.0175 value. Is that from something I wrote? My eyesight is no better than average. This isn’t about eyesight, or what I can discern in a print. What I’m trying to get at is that there is not a single circle of confusion value, no matter how strict or loose, that can create charts that work for every landscape. And that’s because the best possible result you can achieve — the smallest possible circle of confusion — varies wildly from landscape to landscape, depending upon how close the nearest subject is. If everything in the landscape is at infinity, you will necessarily be able to achieve sharper results than if the foreground is one foot from your lens. So, you essentially have to use a hyperfocal distance chart of different strictness depending upon the landscape. Is there a way to find the best possible strictness, given inputs about how close your closest object is? Indeed there is; it isn’t some hidden knowledge that we haven’t yet discovered, or mathematically can’t discover. It’s a simple formula that’s found at all the resources I’ve been linking. (The best explanation of the formula and its derivation comes from George Duovos’s section “Calculating the Sharpest Aperture,” and, although it is easy to understand, it’s infinitely better to read it somewhere that is formatted properly rather than via an unformatted comment like this. Otherwise, I would attach it here.)
If you want to think about it like this, what I’ve been describing so far is essentially to plug in the value of your nearest foreground object into the formula at each of those resources, as well as your focal length. Once you use them to calculate the minimum achievable circle of confusion, you create a brand new hyperfocal distance chart using that value rather than 0.03 (or any other value). Of course, that’s far more work than the steps I outlined in the “sharpest aperture” article, but it will provide the same result in the end, and you can use that method if you prefer. I really hope that this part finally makes sense, because I don’t think I have an easier way to describe it.
To your final two paragraphs: You are right that diffraction decreases overall image sharpness, of course. The goal of this formula isn’t to capture the sharpest image at the point of focus — which would always occur at the lens’s sharpest aperture — but to maximize foreground and background sharpness, taking into account both diffraction and depth of field, so that the entire image is as sharp as possible from front to back. Going back to the 24mm example in our prior comment: If you can name a single focusing distance and aperture that gives you a smaller circle of confusion at both the closest and farthest points than 0.026mm, and back it up mathematically, I welcome it, and I will readily admit whatever mistake I have made.
You’re 95% of the way toward understanding this already, which is why I’m spending so much time on this discussion. I think you’re at the point where you know enough to make that final leap and realize that we do, indeed, already have a “one-size-fits-all” solution, if that’s what you like to call it. The one-size-fits-all solution is to realize that each landscape has a different “best case scenario” minimum circle of confusion, that we are able to calculate this minimum circle of confusion accurately, and that we can create hyperfocal distance charts specific to a given landscape once we know that value — the results, of course, being sharper than what you could achieve by following a 0.03mm chart (or any other value) for every single landscape. (And that if you don’t want to go through the trouble of calculating a separate hyperfocal chart for every scene, you can follow the process I outlined in my “sharpest aperture” article instead, to get the exact same result.)
In your link to LargeFormatPhotography that you have urged me to read, check the section headed ‘What is an acceptable circle of confusion?’ It’s right near the top, appropriately enough, yet this fundamental consideration does not appear in your articles. Then you should know where I got 0.0175mm as your proposed new primary CoC, and will hopefully understand that DoF and HFD tables are not about maximum sharpness per se, but about maximum perceived sharpness, according to the rules of print size and viewing distance.
And following on from that, you will also see how DoF and HFD calculations are absolutely right and accurate, and that there is no point in attempting to modify them because you cannot improve the perceived maximum sharpness without better eyesight, or by breaking the rule of viewing distance – you simply won’t see it any other way.
What you’re trying to do appears to be something different, and plays by different rules. In which case you shouldn’t start by rubbishing DoF and HFD charts, because a) that would be a mistake, and b) will run you straight into robust resistance from folks like me, many other posters on here, and I see yet more on DPReview (where I picked up this link).
Start with a blank sheet, explain what you are trying to demonstrate and the benefits, and keep references to conventional DoF and HFD calcs separate. Also, one of the great things about photographic theory is that if anything is worth doing, then you can readily see the benefits visually in an image. Equally, if you can’t, then it probably isn’t worth spending too much time on. And good luck with it – you say this theory has been around since 1996, perhaps it’ll catch on soon.
FWIW, I think you are just tinkering around the edges, tweaking a few things (in ways that experienced photographer already do anyway) and claiming to have found hidden treasure. The changes you’re suggesting to established practise are relatively minor, will be hardly noticeable, and are almost within the normal margin of error in practical terms. Because, despite what the numbers suggest, depth of field isn’t a very exact science and sharpness doesn’t suddenly disappear over a few inches in landscape photography. It fades very gradually and you can be quite a long way off optimum settings before anything becomes obviously wrong. With that in mind, you don’t need a vast table of data to work from, and inside the lens cap of my 17-40mm super-wide I have just a handful of number jotted down. That is, hyperfocal distance settings for 17mm, 24mm and 35mm, at apertures of f/5.6, f/11 and f/22. In practise, that covers everything I need to know for landscape working, and I can estimate intermediate settings accurately enough, and maybe tweak them a bit according to circumstance, confirmed by chimping.
If I run out of ideas there, then the solution is not to grab yet another set of tables, but to change the game and use something that will make a real and substantial difference to both DoF and overall sharpness – a tilt & shift lens.
Hi Richard, I think that we are essentially on the same page, and that we might only disagree about semantics, but I’d like to confirm.
Before that, though, I will briefly point out that in the specific example I provided, it’s true that the differences are only about 13%. But, as the George Duovos link demonstrates, you’ll frequently run into cases where the recommended aperture via this method, and the recommended aperture via a 0.03mm circle of confusion chart, are two or three stops removed from one another. In cases like that, you lose 31% of your maximum sharpness (if two stops different) and 50% of your maximum sharpness (if three stops different) by following the 0.03mm chart. So, these certainly aren’t within the “normal margin of error.” However, I see where you’re coming from, and I don’t really dispute it; it is true that in many cases, the differences between this method and just defaulting to 0.03mm CoC aren’t mindblowingly different. You’re also quite right about tilt-shift lenses, of course, although that is more of a specialty fix.
So, I’ll revisit this scenario again: Say that the closest object in your scene is four feet away, so you focus at eight feet. If you agree with me that the sharpest aperture setting at this point, assuming that you want the foreground and background both to be as sharp as possible, is f/13.4, then we are on totally the same page, and any disagreements are purely semantic (i.e., whether hyperfocal distance charts are “wrong,” given that they still provide useful information).
If, however, you disagree with that statement, please let me know what a sharper aperture would be. If you cannot provide one, and demonstrate mathematically that its circle of confusion at the closest and farthest distances are both smaller than 0.026mm, then we are done here, as everything I say beyond this point will fall on deaf ears.
I don’t dispute the accuracy of your figures, not that I’ve checked them but I’m sure they work as you say. Your suggested method and tables are not for me, as I’m not convinced they do anything significantly better than a slight tweak to normal HFD settings achieves, and I like to keep things simple, and separate. I use HFD with the double-distance method, a focus tweak front or rear, maybe an f/number adjustment and a chimp to check. I understand diffraction, and I know the optimum apertures of my lenses. If I was a keen landscaper seeking the sharpest possible result, I wouldn’t bother with any of your minor tinkering and simply use a tilt & shift lens that takes everything up to another level. If you’re looking for a game-changer, that knocks your new charts out of the park.
What concerns me though, is your dismissal of HFD and DoF theory and practice, as per your attention-grabbing headline. You’ve had a lot of grief over that, but you brought it on yourself and I’m still not sure you know why. Perhaps you do, but an acknowledgement is conspicuous by its absence. The principles that underpin both HFD and DoF are 100% correct according to the standards specified, they work, they actually cannot be improved, and they’ve been universally used and accepted for, I don’t know, since forever. They’ve stood the test of time because they relate to our standard of vision, not to cameras and lenses per se, and they are referenced throughout the graphic arts industry. You will be aware of the 300dpi standard for all aspects of high quality print media; if you take the primary circle of confusion used in the HFD and DoF formulae of 0.2mm, that translates to 254dpi which in practical terms is effectively the same thing.
HFD and DoF formulae are not intended to give the sharpest possible result, which is what you are after, and they never have been. They’re intended to deliver maximum perceived detail visible by the human eye, at normal viewing distances – and it’s human visual acuity that sets the limit, regardless of ever sharper lenses or increased pixel counts – or fancy new formulae. The HFD and DoF calculations are independent of all that, which is why they’re as relevant today as 100 years ago and stand as a universal reference constant that we can always rely on.
It’s true that it’s an attention-grabbing headline, more so than necessary. That’s my bad. I’m not a good headline writer, but, in hindsight, something like “How to capture a sharper image than what hyperfocal distance charts recommend” would be more accurate, and potentially less inflammatory.
I don’t mean to dismiss hyperfocal distance or the idea of depth of field. Hyperfocal distance charts and calculators do exactly what they are meant to do — provide you with the settings that will result in a circle of confusion with a size of exactly 0.03mm in the background (or whatever other, smaller or larger, circle of confusion you prefer for a given image/output size). For their stated purpose, they are totally accurate, and not “wrong” at all.
You’re also right that not everyone will need the added sharpness of calculating the optimal aperture. If you’re printing at small enough sizes, it’s physically impossible to see a difference between — for instance — 0.02mm of background CoC versus 0.01mm CoC. That’s the case for the example you just presented, where, in an 8×10 print (which I assume is the print size you’re referring to), you’ll end up with an effective 254dpi in terms of background sharpness if your hyperfocal distance chart is calibrated to a 0.02mm CoC. Personally, as someone who doesn’t have magnifying glasses for eyes, I’d say that’s a pretty remarkably detailed result :)
And, if you change around your CoC value depending upon your output size, that’s even better. If you print 16×20, for example, you might follow a hyperfocal distance chart based upon a 0.01mm CoC, and so on. (To be fully accurate — you might do so if you’re prioritizing background sharpness more than foreground sharpness, because, as you said, these charts don’t have any relevance toward how sharp your foreground appears, and the wide aperture and far focusing distance required by a 0.01mm CoC chart could potentially leave you with a somewhat blurry foreground.)
The techniques I mentioned are only accurate if you’re prioritizing foreground and background sharpness equally, and you want the sharpest possible result in case you ever make especially large prints. It’s entirely true that they make no claim about whether you can actually see that increase in sharpness :). And, though I’ll point out that there isn’t an instance where a traditional hyperfocal distance chart provides a sharper front-to-back image than the method described here (regardless of print size, output medium, scene depth, individual eye acuity, viewing distance, etc.), getting the technically “maximum” sharpness often isn’t necessary, especially if the differences aren’t even visible in your largest potential output. If I knew that my largest output size was going to be an 8×10 print, for example (just an arbitrary choice), there would be no need to take the time to maximize sharpness in the original file, since any differences (especially in cases where the “optimal” method is only 13% sharper) would be functionally invisible. Indeed, this isn’t something I’ve talked about enough so far, and I’m sorry that it took me so long to realize that this has been the crux of your comments.
I hope you can see from the sheer number of hours that I have spent replying to people’s questions that I never intended for this discussion to become so chaotic or unmanageable. My initial goal was to provide photographers with a technique I find useful to capture the sharpest possible images, if they aren’t satisfied with the sharpness from following hyperfocal distance charts for their particular uses. And although the mathematical underpinnings of this technique are solid, that doesn’t mean this is the proper method for everyone. If you’re already using hyperfocal distance charts without issue — and it seems that you, personally, have found a technique that works excellently for you — that is just fine, and I wouldn’t expect you to switch.
No matter which technique you use, I wish you the best of luck with it. I think that both of our goals are exactly the same: help beginning/intermediate photographers learn more about how to capture the images they have in mind. For that, I appreciate all the time you’ve taken to spell out your perspective, and hopefully readers in the future will find this back-and-forth to be useful in shaping the way they approach this topic.
Cheers Spencer. Thanks for the debate, and your responses under fire! BR, Richard :)
Sorry, Spencer, but your definition of HFD just does not ring correct:
It’s the closest point to your camera that you can focus, while still ending up with an acceptably sharp background.
It has more to do with the focus at infinity than the close focus::
The hyperfocal distance is the closest distance at which a lens can be focused while keeping objects at infinity acceptably sharp. When the lens is focused at this distance, all objects at distances from half of the hyperfocal distance out to infinity will be acceptably sharp. (From wiki)
I have no issue with that definition. It’s completely accurate. The issue is that hyperfocal distance charts simply don’t tell you the optimal spot to focus; they tell you the spot to focus that will result in exactly identical background sharpness for every single photo, not taking the foreground into account whatsoever. Whether your foreground is touching your lens, or is a thousand meters away, those charts will recommend the same focusing distance.
They aren’t wrong, by any means, if your goal is to find exactly the spot to focus that results in a 0.03mm circle of confusion in the background. But, for most landscape photographers, that isn’t their goal. Instead, they want to find the focusing distance that equalizes the foreground and background circles of confusion. Then, they want to find the aperture that minimizes the circles of confusion, taking both diffraction and depth of field into account. By a wide margin, that isn’t something hyperfocal distance charts tell you.
No, your definition is wrong. And because your definition is wrong, the whole premise of your article is suspect. Hyperfocal distance does not tell you about the foreground because it is not designed for the foreground. It tells you where the closest focal point has to be to keep the infinity in focus. All points halfway from the focal point and the lens will be in acceptable focus, the rest being not in focus. You have to understand the basic concept before you can claim something is wrong!
Once again, we’re on the same page.
The definition of hyperfocal distance is the closest point you can focus and keep infinity acceptably sharp (not “in focus” as you said, but that’s a minor quibble).
And, yes, the region at half that distance will also be acceptably sharp, because its sharpness will equal the sharpness in the background.
What sort of dispute do you have with my definition? It seems to be the same as yours.
“You have to understand the basic concept before you can claim something is wrong!”
You also have to understand the basic concept before you can claim someone is wrong!
Have you considered the possibility that the person who is wrong, in this case, is you. You are making the claim that the author of the article is wrong, therefore, you own the burden of proof for your claim.
“That which can be asserted without evidence, can also be dismissed without evidence.
— Christopher Hitchens”
You wrote “[Hyperfocal distance] tells you where the closest focal point has to be to keep the infinity in focus.”
Which is, of course, abject nonsense. When the lens is focussed to the hyperfocal distance then the only objects in the scene which are, by definition, “in focus” are the objects in the scene residing at precisely the infinitely-thin plane of focus: the hyperfocal distance away from the camera. All other objects in the scene are, by definition, out of focus.
The main problem I see here is not so much with the article (although there are elements which are debatable), but rather the headline.
Everything discussed here relates almost entirely to landscapes, which is fine, but hyperfocal charts are probably of least use in that genre! Given the time to set up, you could double check everything and carefully consider your options. Having an idea of hyperfocal distance is most useful when everything is moving, you can’t be sure where your subject will be, and you don’t have the time to be certain of focus. Certainly these days, with fast AF, it’s less of an issue, but not so many years ago it gave you the best chance of ‘keepers’ if using hyperfocal settings.
I think the headline should be: “Why Hyperfocal Distance Charts Are Not Perfect For Landscape Photography”
“Everything discussed here relates almost entirely to landscapes, which is fine, but hyperfocal charts are probably of least use in that genre! … Certainly these days, with fast AF, it’s less of an issue, but not so many years ago it gave you the best chance of ‘keepers’ if using hyperfocal settings.”
So, please explain how hyperfocal distance charts aid other genres, such as portrait photography and macro photography. [Hint: They are useless!]
I didn’t suggest them for portraits, or macro. In fact, I specifically mentioned motion.
Fair points. It is true that this article assumes you are actually able to set the recommended aperture (which will generally be smaller than what hyperfocal distance recommends) in the field without issue. Although that usually is true, it’s not always the case.
For nighttime landscape photography, for example, you’ll be at f/2.8 (or even wider) quite frequently. If your main goal in that scenario is to capture a sharp night sky, focusing twice as far as your foreground obviously won’t work (given that your foreground is nearby). That isn’t a failure of the double-the-distance method, though; even in this scenario, it will equalize sharpness between the foreground and the background. Instead, it’s a situation where you are prioritizing background sharpness rather than foreground sharpness. You inherently have to make compromises in order to capture your subject at any degree of sharpness.
Luckily, for the vast majority of landscape/architectural photography, this isn’t the case. Instead, you’ll have a lot of flexibility as to what aperture you can use. When you do have that flexibility, combined with the goal of ensuring equal sharpness between the foreground and background, it is quite true that the DTD method is the optimal solution (combined with selecting the sharpest possible aperture).
I’m fairly sure that we’re on the same page for all of this, but let me know if there’s anything else you’d like to add.
Subject motion occurs in the genres of: street photography; stage photography; portrait photography; wildlife photography; seascapes; cityscapes; and almost every genre other than still life photography. Photographing a bunch of dried flowers in a vase on an indoor table is one of the few genres that does not include much motion.
Please provide practical examples of your claim that “Everything discussed here relates almost entirely to landscapes, which is fine, but hyperfocal charts are probably of least use in that genre! … Certainly these days, with fast AF, it’s less of an issue, but not so many years ago it gave you the best chance of ‘keepers’ if using hyperfocal settings.”
Spencer
I welcome your post, but I think you are ‘wrong’. I know that is a blunt statement, but let me explain how I see things.
If we ignore diffraction for now and assume a simple lens and make a few other assumptions, the HFD is simply (FL^2)/(N.C), where FL is the focal length, N the aperture number and C our selected CoC or defocus, blur criterion, ie 30 microns.
Noting that 30 microns is a just acceptable (FF) criterion and that it is not sensible to focus at (sic) infinity, ie when the blur is zero, as blurs less than, say, two sensor pixels are rather meaningless; we can put a sensible ‘limit’ of ‘far field focusing’ of xHFD, where x is between 2-3. That is, if we focus at, say, 2xHFD, the defocus blur at infinity will be the CoC/2. At a focus 3xHFD, the infinity blur will be CoC/3.
If x is less than 1, ie we are focusing shorter than the HFD, then the near and far depths of field are simply, near DoF = xHFD/(1+x) and the far is xHFD/(1-x).
If focusing shorter than the HFD, a useful ‘trick’ is to simply remember the HFD and use x as a fraction x =1/n, where n is an integer greater than 1, which will give you the following information that is very easy to calculate in your head, eg for n = 3, ie we focused at HFD/3:
– point of focus = HFD/3 or generally HFD/n
– near DoF = HFD/4 or HFD/(n+1)
– far DoF = HFD/2 of HFD/(n-1)
The bottom line being that if you focus shorter that the HFD, than the far DoF will never reach infinity and (sic) meet the CoC. The above also clearly shows that as x approaches 0, ie highly macro, the DoF becomes small and symmetric. For example, if we (could) focus at, say, HFD/10, the depth of field is HFD/9 – HFD/11 = 2HFD/99 or about one 50th of the HFD.
As a landscape photographer, I use the HFD, and will always focus ‘long’, ie I longer than the HFD, to achieve the minimum infinity blur, ie x = 2 to 3. Thus if I know I want an infinity defocus blur of 15 microns, I will focus around (sic) twice the HFD that is given by a 30 microns CoC.
If required, I will then focus stack at HFD/(2n-1), ie HFD, HFD/3, HFD/5, HFD/7 etc if I want to cover the near field.
BTW the number of focus brackets that is required, ie n, can be estimated from HFD/(2x), rounded up: for x less than zero. For example, if the HFD was 10m and I wanted to include a point of interest that was at 1m, then I would need to take 10/2 focus brackets, ie 5, at HFD, HFD/3, HFD/5, HFD/7 and at HFD/9: ie at 10, 10/3, 10/5, 10/7, 10/9 metres.
I am in the process of writing about this approach, ie ‘just’ remembering the HFD, on my blog at photography.grayheron.net. I should have a post up this weekend.
Garry,
Suppose we use a 100 mm macro lens at f/8, at a focus distance = hyperfocal distance (H) / 200.
H = 41.8 m, using CoC=0.03 mm
object distance from the front principle node/plane of the lens ≈ 0.2088 m
object-to-image magnification ratio ≈ 0.9188
DoF near ≈ 0.54 mm (in front of the object)
Dof far ≈ 0.55 mm (behind the object)
Dof total ≈ 1.09 mm
Using your statement “If x is less than 1, ie we are focusing shorter than the HFD, then the near and far depths of field are simply, near DoF = xHFD/(1+x) and the far is xHFD/(1-x).
…
– point of focus = HFD/3 or generally HFD/n
– near DoF = HFD/4 or HFD/(n+1)
– far DoF = HFD/2 of HFD/(n-1)” we get:
x = 1/200 = 0.005
DoF near = x·H/(1 + x) ≈ 207.8 mm
DoF far = x·H/(1 − x) ≈ 209.9 mm
DoF total ≈ 417.7 mm
n = 1/x = 200
DoF near = H/(n + 1) = H/(200+1) ≈ 207.8 mm
DoF far = H/(n − 1) = H/(200−1) ≈ 209.9 mm
DoF total ≈ 417.7 mm
It seems to me that either: my calculations are hopelessly incorrect; or your formulae are hopelessly incorrect. My experience with macro photography makes me very much doubt that the depth of field is circa half a metre when using f/8, on FX format, at a magnification ratio close to unity.
By all means, explain to the readers and to myself, my errors.
Pete
Should have said the equations are not good for macro, as you need to throw in a magnification term.
Also, DoFs are absolute distances from the camera.
Thus, using your example:
n = 1/x = 200
DoF near = H/(n + 1) = H/(200+1) ≈ 207.8 mm
DoF far = H/(n − 1) = H/(200−1) ≈ 209.9 mm
DoF total ≈ 209.9 – 207.8 ≈ 2mm, but we have ignored mag.
Cheers
Garry
BTW have just posted this: http://photography.grayheron.net/2017/09/unlocking-power-of-hyperfocal-distance.html
Garry, Thank you for your explanation.
There most certainly are many circles of confusion regarding the terms “DoF near” and “DoF far” and their ilk :-)
One of my very few specialities in the fields of mathematics and physics is optical physics, therefore I have an aversion to terms and phrases which do not have internationally-agreed names.
I used the correct trigonometric rectilinear equations, not any of the simplified equations, in my example therefore I’m pleased to find that your equations [when we’ve eventually managed to figure out how to interpret your writing!] are within a factor of 2 of the correct value for macro magnification ratios approaching unity.
In my opinion, based on my decades of experience, your equations seem to be fit for their purpose, but your chosen nomenclature — e.g. “HFD” — and your lack of conformance with ISO standards for technical writing do, unfortunately, make you appear to be nothing other than a random Internet poster/blogger. Absolutely no offence intended, I think your many hours of work, perhaps, deserves much better recognition.
I shall not post any comments on your blog/article because Nasim’s Photography Life website is the only photography website with which I’m willing to freely share my knowledge of photography-related mathematics and science.
Your article would’ve been more accurate, and more useful to photographers, if under its title Unlocking the power of the Hyperfocal Distance you provided little more than just links to the relevant, excellent, accurate, succinct, and well-illustrated articles on Photography Life.
It is very difficult to refrain from drawing the obvious conclusion that your comments on Spencer’s article are nothing other than self-promotional advertising.
Pete
This particular post has generated a lot of response. I wasn’t going to reply to you, as I thought there was enough ‘bad blood’ already. However, I would like to say a couple of things.
First, I enjoy reading Photographylife and when I believe I have something to say, I will add my thoughts.
Second, I write my blog for my personal, carthatic reasons, and for friends. I certainly have no desire to ‘publicize’ my personal thoughts on photography.
Thirdly, I admit I’m not in the same league in photography as obvious experts, such as yourself.
Fourth, I try and use accepted terms such as near and far depth of field etc, and didn’t know there was ISO standards to follow. I’ll be more careful in the future.
Bottom line: my post here was my humble attempt to help others get more out of the hyperfocal distance (HFD) approach, and ‘correct’ what I believed to be a ‘confusion’ in the original post.
All the best.
Garry,
Thanks for your reply. It seems to me that you’ve put much time and effort into your article therefore it’s only fair that I should put more time and effort into my replies to you. I’ve written the following just in case it contains a snippet of information that is useful to you. Please do not take it as personal criticism…
A numerical value must be separated from its units by one space character (20 mm), and it’s a good idea to use a no-break space character because this avoids the possibility of an end-of-line wrap splitting the value and its units.
When writing equations, variable are italicized and constants are not, e.g.,
E = m·c², note that I’ve used the dot operator for clarity, which wasn’t necessary here, but it’s very useful when two or more variables are multiplied.
In an earlier comment you wrote “the HFD is simply (FL^2)/(N.C)”, and here are my suggestions for improvement. Hyperfocal distance is usually denoted by the symbol H, focal length by the symbol f, circle of confusion c [sometimes C], and f-number N [or k in some documents], and the correct formula is:
H = f² / (N·c) + f
Obviously, in the comments section of an article it’s fine to write
H = f^2/(Nc) + f, but using the well-known symbols, especially the correct formula, informs the readers that the author of the comment probably does know at least the basics of the subject. This is particularly important when disagreeing with the author of the article on which we are commenting.
Some of the above may appear to be semantic nitpicking, but mathematics and science would be useless if they were not semantically correct, unambiguous, precise, and accurate.
My final suggestion is to treat the authors of Photography Life articles with respect because they really do know what they are talking about. If we disagree with something they’ve written it’s nearly always because we haven’t understood it correctly or because we have been misguided by the huge volumes of inaccurate information which reside on the World Wide Web.
If you have any technical questions relating to Spencer’s article, I’m more than happy to answer them.
Best wishes,
Pete
Pete
I really appreciate your reply, especially the emphasis on ‘type setting’. Your comments mean I will follow your guidance on my blog.
BTW I know the hyperfocal distance formula I am using is an approximation, i.e. I ignored the additional focal length term, but this is reasonable for landscape work. Hence, ‘my’ equations are limited to non-macro.
I’m also aware I ignore pupil magnification, but once again, for most landscape cases I believe that this is not unreasonable.
In my Lua scripts, that do auto landscape bracketing in-camera, via Magic Lantern, I use the ‘correct’ equations for depth of field, albeit ignoring pupil mag.
The approach I have explored, i.e. ‘just’ using H to calculate focus stacking distances, is a handy approach. That is remember one distance and simply multiple or divide by integers. This is also a reasonable way to estimate near and far depth of fields at any distance shorter than H, i.e. when doing portriat work, or where you wish to isolate the background. All you need to do is know H and work our the fraction of H you are focused at.
Once again, thanks for taking the time to reply to me and point me in the ‘right direction’. I much appreciate your response.
Cheers
Garry
Garry,
Again, just in case the following is useful to you…
I agree with you that ignoring the focal length term and ignoring pupil magnification seems to be reasonable for the majority of landscape photography, because the simplified formulae are, I think, ‘reasonable approximations’ for small format cameras [135 (FX) format is a small format]. But, these simplifications grow increasingly unreasonable as the format size increases towards large format and beyond. Therefore, I would write something along the lines of:
H = f² / (N·c) + f, however, when using small format cameras for typical landscape photography, we can simply this to:
H ≈ f² / (N·c)
Note the change from “=” [equals] to “≈” [almost equal to / approximately equal to].
Approximated formulae were very useful before the advent of computers, because everything had to be handwritten on paper, and formulae had to be painstakingly evaluated with the aid of books of mathematical tables. Now that we have computing devices for both writing documents, and the evaluation of formulae, we can no longer logically justify the use of approximations instead of the correct formulae. But, we can logically justify: deriving the exact formulae for a specific application, then simplifying it as the final step — definitely not in the initial step or during the intermediate steps.
I’m not for one moment suggesting that you should do it; I’m suggesting it just in case you find the process enjoyable and satisfying. There is — very sadly in this 21st Century — another reason why it could be useful for you to do it: your writing will provide less potential fodder to random Internet trolls.
Best wishes,
Pete
Pete
A final thank you, for your time.
Also, to make the H approach work down to macro, all you need to do is account for the focal length, f.
Thus Dnear = s (H – f) / (H + s – 2f) and Dfar = s (H – f) / (H – s).
Where s is the focus distance; and Dnear and Dfar are the distances to the DoF limits.
BTW I don’t know how to get italic text in this reply.
Cheers
Garry
Garry,
Your equations for Dnear and Dfar are correct, provided that the subject distance s is the distance from the front principal point of the lens, not the distance from the sensor.
I’m sure you know that, but I’ve mentioned it because it’s a frequent cause of bewilderment. E.g., I’ve seen people engage in long debates as to why the stated closest focus distance of a lens does not correspond to its stated maximum reproduction ratio. The reason for this discrepancy is threefold using modern lenses:
1. Their stated focus distance is from the sensor;
2. their pupil magnification ratio is unlikely to be unity [Note 1];
3. internal focus lenses change their focal length f with focus distance [Note 2] and their stated f applies only when the lens is focussed at infinity.
Note 1: A telephoto lens is, by definition, a lens which has a pupil magnification ratio less than unity; an inverted telephoto [aka: retrofocus] lens has a pupil magnification ratio greater than unity.
Note 2: Technically, they change their optical power with focus distance, whereas focal length is just the thin lens model equivalent of their optical power.
PS: Don’t bother with italics in comments because it’s far too easy to make a mistake which renders the comment almost unreadable. I’ve made mistakes with text markup tags so many times that I ended up writing a parser to check my comments before I post them.
Pete
I was aware of a difference, but have not been able to find a definitive source, that says this.
Wiki doesn’t seem to explicity reference the sensor vs the principal point.
Do you know of a link, ie for my education.
Cheers
Garry
Garry, I’ve spent, at the very least, a few thousand hours of my life attempting to find truly definitive, freely-available, scientifically peer reviewed, sources of accurate information on photography-related optical physics. I managed to find a few extremely good sources of freely-available information, but even these failed the acid test of the proper meaning of the term “definitive”, because they each contain at least one approximation of the exact mathematically-correct equations.
So, I am unable to provide you with freely-available, abjectly definitive, sources of thoroughly peer-reviewed documents. Very sadly, the truly excellent source of information, freely given to us by Paul van Walree via his website “toothwalker.org”, is currently unavailable. It was available on 2017-02-03 when I pointed the readers of Photography Life to Paul’s exquisitely explained and illustrated article entitled Depth of field. I guess Paul’s website has been archived by at least one of the ‘Internet archive’ repositories, but I’ve been unable to find a link.
Other excellent resources of science- and evidence-based information are, of course, the plethora of technical documents published by Hasselblad, Nikon, Olympus, Panasonic, Schneider, Zeiss, et al.
Garry, I managed to find an archive:
Derivation of the DOF equations
https://web.archive.org/web/20160730041911/http://toothwalker.org:80/optics/dofderivation.html
The Wikipedia article Hyperfocal distance shows that H is the distance from the centre of the thin lens, not the sensor. The applies to all object space distances.
https://en.m.wikipedia.org/wiki/Hyperfocal_distance
Pete
I guess I’ll stick with my ‘definitive treatise’, ie The INs and OUTs of FOCUS: An Alternative Way to Estimate Depth-of-Field and Sharpness in the Photographic Image by Harold M. Merklinger.
Page 15 does it for me ;-)
Cheers
Garry
Pete
Once again thanks for the links. I remember reading Toothwalker in the passed.
As life is an approximation [;-)], and the DoF equations assume a simple lens, ignore pupil mag, I will make the ‘reasonable’ assumption that the sensor is about (sic) a focal length away from the principal plan of the lens. Don’t shoot me!
After all, for most/all landscape work the sensor plans is essentially the principal plane, ie this only becomes an ‘issue’ for close in macro work. Plus I tend to use WA lenses, ie focal lengths insignificant to relative to far field distances.
Cheers
Garry
Garry,
Thank you very much indeed for our interesting, mutually-informative, and on-topic discussion. I think we’ve been very lucky to escape having our discussion interrupted/subverted by trolls.
Best wishes,
Pete
Perfect. A general rule even I will be able to recall 8 hours from now. Thanks!
I couldn’t disagree more.
Depth of field charts are not “spectacularly” wrong. The charts are calculated using a circle of confusion that provides a perceived degree of acceptable sharpness, that certainly works well for most applications. The word perceived is important in that objects both in front of and behind the focusing point (plane) are always lacking perfect sharpness. It is simply that the the lack of sharpness is too slight to be perceptible. It doesn’t matter how you determine focusing distance, objects outside the focusing plane are never in focus. What matters is whether that lack of sharpness is perceptible or not. Depth of field scales are calculated specifically for different film and digital sensor sizes so your argument that the charts are somehow outdated is not correct. If your needs are really exceptional, and you truly can see a lessening of sharpness at infinity, lets say, compared to the hyper-focal distance, then you can simply make a small compensating adjustment to the focusing distance (or aperture). How to do so, becomes obvious if you have a good understanding of depth of field. Or you can compensate by adjusting input settings on a DOF calculator… type in a smaller CoC (circle of confusion). Simple.
Sorry, but the double-the-distance idea is useless for landscape photography. Determining the closest object you want in focus, and then doubling the distance does not take into account the most most distant object you also want in focus. You aren’t even factoring in the changing depth of field when using different apertures. If the foreground subject is 1 meter away and you focus on 2 meters will infinity be in focus? Impossible to know. If you need a background object 20 meters away in focus, which aperture to use while keeping 1 meter also in focus? Impossible to know, unless you use a depth of field chart or calculator.
I suggest to the readers that you download a depth of field calculator to your smartphone. It is a superb tool that provides a wealth of information furthering your understanding of depth of field principles. Regards
Is your goal to equalize sharpness (i.e., circle of confusion size) between the foreground and background? That’s the case for most landscape photographers, but it sounds like that’s not what you’re saying in this comment.
If that is your goal, doubling the distance most assuredly does work. It undeniably equalizes the circles of confusion between your nearest and farthest objects. The only issue, then, is to pick the proper aperture that minimizes both (identical) circles of confusion. That’s not something traditional hyperfocal distance charts or calculators can tell you. If you’re interested in exact accuracy, you should read this article: https://photographylife.com/how-to-choose-the-sharpest-aperture, as well as the optical science underpinnings behind it: http://www.georgedouvos.com/douvos/OptimumCS-Pro_Optical_Science.html
And if you like apps, I would suggest one called OptimumCS. It’s a paid app ($5), and I have no affiliation or relation to it in any way. However, it also happens to be the only one out there (unless someone recently released a new one) that calculates all of this accurately.
I’ll end by answering a question you posed: “If the foreground subject is 1 meter away and you focus on 2 meters will infinity be in focus? Impossible to know.”
It is, in fact, not impossible to know. (First, I’ll assume that you meant “sharp” rather than “in focus,” because only a single plane can be in focus at a given time.) In fact, you can know with utter certainty that infinity is equally as sharp as the region at 1 meter away. They have identical circles of confusion. That’s the very definition of depth of field in the first place; it’s even in the intro to the Wikipedia article on hyperfocal distance.
Whether that 1-meter region and that region at infinity are as sharp as possible, though, depends upon the aperture that you use. There is an optimal aperture in this scenario for every given focal length, and the formula to calculate it has nothing to do with hyperfocal distance charts. Once again, check out both links I added, and you’ll find all the optical science that went behind determining the correct answer. (Just to give a couple examples: In this scenario, with a 14mm lens, the sharpest aperture is f/8.6. With a 20mm lens, the sharpest aperture is f/12.4. And so on.)
For anyone not willing to actually calculate the CoC diameter that should be used for DoF calculations – a CoC diameter that takes into account the anticipated enlargement factor and your desired resolution in the final image (expressed in lp/mm) for an anticipated viewing distance…
For a quick and easy exploitation of the principals of hyperfocal focusing…
In addition to focusing at twice the distance to the nearest subject in your frame…
Consider setting your camera to Shutter Priority mode, at the highest ISO setting you are willing to suffer (in terms of sensor noise), then always select the Shutter Speed that is only just fast enough to freeze both camera and subject motion.
This choice of the slowest viable shutter speed comes quickly with experience, but the good news here is that your Shutter Priority mode camera will then always be using the smallest Aperture possible for the prevailing light and motion, thus always yielding the greatest possible Depth of Field.
No CoC selection or DoF calculations are required, doing this, but the downside is that you’ll never know if the DoF was sufficient until you examine your final image, after enlargement, at the viewing distance you prefer. And… if your enlargement factors are unusually high (as can happen with tiny sensors having high pixel counts) or your viewing distances are unusually close relative to the final image diagonal), the smallest possible aperture for the prevailing light and motion (had when selecting the slowest viable shutter speed, especially in combination with a high ISO) may end up causing visible softening across the entire image, due to diffraction.
Convenience seldom comes without compromise. but for those who prefer to avoid doing calculations in the field or referring to pre-calculated DoF tables (even those that have been prepared with smartly chosen CoC diameters), focusing at twice the Near distance and choosing the slowest viable shutter speed in Shutter Priority mode, can yield pretty satisfying results for all but the most extreme combinations of enlargement factor and final image viewing distance.
I don’t understand how I determine the correct stop. I’m using a Canon 18-135mm lens with a T6i bpdy.
Once you’re focused at double the distance to your closest object, there are a few ways to find a good aperture. The first one is just to go based off your personal experience; does the landscape in front of you look like it will require a ton of depth of field (like, if the closest object is one foot away from you, and you’re focused at two feet)? Or, will the landscape call for less total depth of field (like, if you’re at an overlook, and everything is in the distance)?
If the landscape needs a lot of depth of field, use something like f/11 or f/16. If the landscape doesn’t need as much depth of field, you can use something like f/5.6 or f/8 instead. Particularly when you’re starting out, it’s hard to go wrong just by setting f/8 or f/11 if you aren’t sure.
When you get to the point that you’re fairly comfortable with how aperture works, you might take a stab at reading my article on how to choose the sharpest aperture (https://photographylife.com/how-to-choose-the-sharpest-aperture). This one tells you the mathematically optimal aperture for a given focal length and focusing point, assuming that you want your foreground and background both to be as sharp as possible. However, it’s certainly a more advanced article, and it isn’t the type of thing that I recommend worrying about yet. You’ll get more value by simply going out in the field and experimenting with different apertures, keeping in mind the rough range of f-stops that works best for a given landscape.
I hope this helps!
Very good article and very well written. I’ve never had any luck with the hyperfocal distance tables so this really piqued my interest.
I’m having trouble understanding which fstop you use in any given situation. Everything else makes perfect sense but I’m not sure which fstop to use for a specific lens length and focusing distance. I’m using a Canon 18-135mm lens with a Canon T6i body.
Thank you, Mike! Did you see my comment #48.1? I tried to answer your question, but I might not have provided the level of detail that you wanted. Is there anything you’d like me to add? Best of luck!
Professional landscape photographers, as opposed to casual snapshot nature photographers, want everything to be in focus (acceptably sharp). They use depth of field charts to determine which aperture to use; the aperture that provides the needed depth of field for the particular scene they are photographing and the particular focal length lens they are using. Saying that DOF charts don’t provide that information is totally incorrect, as that is their primary purpose. The chart you provided as an illustration for your article is a poor example of a DOF chart as it does not provide information on near limit focusing distance, although one could calculate that value from the hyper-focal distance that the chart does provide. A DOF calculator is the preferred tool as it provides focusing distance information when the far limit is something other than infinity.
The double-the-distance method of focusing is okay for the snapshot nature photographer who don’t mind guessing as to whether parts of the scene are going to acceptably sharp or not. Yeah the foreground and background will be equally in or out of focus, however that is not what a professional landscape photographer wants. He or she wants everything to be in focus (acceptably sharp) always. Using the double the distance method also requires that one guesses as to which aperture to use. Certainly not very professional. The method you present should be seen is an alternative, rather than substitute, for determining focusing distance. DOF charts are superior in that they provide precise information that your method does not.
I don’t know why you commented again at the bottom of this article rather than replying to the response I already took the time to write for you above. Could it be that you were unable to find that comment, or is it that you wanted your new comment to appear at the very bottom, in a more prominent position?
Flip your statements, and you’d be accurate. Hyperfocal distance charts are okay for snapshot photographers who don’t mind losing some sharpness every single time. Focusing at double the distance to your closest object, and then using the aperture that minimizes your circles of confusion (found at all the many links I’ve included throughout this comment section, including http://www.georgedouvos.com/douvos/OptimumCS-Pro_Optical_Science.html and https://photographylife.com/how-to-choose-the-sharpest-aperture) will give you the sharpest possible result, and should be used by “professionals.” (I can’t believe you used that term, as you are hiding behind a veil of anonymity, writing on a website filled with people who do earn a living from selling landscape photographs and teaching photography, including people like me and Nasim.)
Say that you’re using a 24mm lens, and the closest object in your scene is six meters away. Given that you want that object and infinity both to be as sharp as possible, where would you focus? What aperture would you use? And what are the sizes of the foreground and background circles of confusion in that scenario?
I can tell you right now that I would focus at 12 meters and set my aperture to f/8.5. That would result in a foreground and background circle of confusion that are both equal to roughly 0.009mm in diameter.
If you can come up with a focusing distance and aperture that result in sharper values than that, you’re welcome to tell me, and I’ll take back everything I’ve written. If you cannot — and I guarantee that you cannot — then we’re done here. I’ve responded to dozens upon dozens of comments, most of them very appreciative, but some (especially the most recent ones) from people who have no desire whatsoever to learn or grow, and only seem satisfied when they argue with others, however uninformed or baseless their opinions may be.
I never mind comments that correct a mistaken viewpoint of mine. In fact, as you’ll know if you ever read my articles on Photography Life, I love them, because my goal is to be as accurate as possible when I write for others. That’s not what this comment section is becoming; it is becoming a hub of people asking the same question, in the same arrogant way, with me providing the same answer to each one, with no one learning anything new. There is nothing more I have to add that I have not said already.
I have faith that the vast majority of our amazing readers understand the realities of this discussion. Our only goal is to provide new tips for those who are still willing to learn about the amazing world of photography. And, hopefully, I will never stop learning along the way myself.
Firstly I apologize for not clicking the reply link instead of making a separate post.
My intent is not to personally insult, however the double the distance method, as it is presented in the article, involves guessing which aperture to use. That is NOT professional. However using the double distance method in combination with using the optimal sharpness chart provided in your other article does away with the guesswork and is a viable alternative. This combined method of determining focusing and optimal aperture is interesting although really not much different than using a DOF calculator combined with intelligent decision making. Landscape photographers who use DOF charts will always strive to use the optimal aperture for sharpness whenever possible, using a smaller than ideal aperture only when the extra depth of field is needed. The hyperfocal distance is also only used only when necessary. Your example of the distant mountain scene not needing the hyperfocal distance provided by a DOF chart is an obvious example.
Saying that hyperfocal distance charts are okay for those who don’t mind losing some sharpness every single time, is a bit absurd. A beginning photographer reading this might think that everything in the photo is somehow less sharp, which of course is not true. If at the extreme near and/or far limit of depth of field there is a perceptible drop off in sharpness (unlikely), then one can compensate by using a smaller CoC. An acceptable circle of confusion limit is considered to be what is undetectable by the human eye. There is no difference in using a method that provides a smaller CoC if that difference is a not a percetible one. All you are doing then is loosing some valuable depth of field.
Unless one is willing to do his or her own mathematical computations the combined method you present is not useful when the far limit is something other than infinity. Please correct me if I am missing something. Going back to my original example… If you need a background object 20 meters away in focus, which aperture to use while also maintaining 1 meter in focus? Easy to determine with a DOF calculator. How do you derive the answer using your method?
Best regards
Shoshone,
Reading back through my earlier reply to you, it’s clear that I have made a mistake — not in content, which I still believe is solid, but in my tone. I have been writing responses to this article, here and on DPreview (which linked to this post earlier), nonstop over the past day or so. I don’t know exactly how many words I’m up to; it’s at least in the 10,000+ range.
But that isn’t an excuse for letting my frustration boil over in the comment above. Your comment was neither the most insidious nor most ignorant of the ones posted here, by a wide margin. Our team here has been deleting the especially cruel and troll-like comments throughout the past few days, and yours were never even close to being among them. So, I apologize for attacking the fact that you simply posted a second comment later on, which is a very normal thing to do, and that my tone was rude throughout the post. There is already so much negativity online, and it was never my intent to add to it. I’m leaving the comment up as-is, but I am embarrassed by many parts of it, and I hope you’ll forgive me.
In terms of content — you are entirely correct that this double-the-distance method is not accurate when your far limit is something closer than infinity. I’m glad you brought that up; I think you’re the first person here to do so. It’s still possible to find the ideal aperture in that scenario, but you have to use the full, unabridged formula:
I would never calculate that in the field. I wouldn’t expect anyone else to do so, either. Since most landscapes have a component at infinity, or far enough away that it is functionally infinity, the good news is that you typically can use the simplified version of that formula, which is this one, where the only difference is that I set “distance to far object” to equal infinity:
I also wouldn’t want to use this formula in the field! Instead, I just use a chart or app that automatically finds this sharpest aperture for me. There aren’t many such apps out there, but the George Duovos ones I mentioned earlier do work. The same is true for the chart in my “choosing your sharpest aperture” article: https://photographylife.com/how-to-choose-the-sharpest-aperture
In practice, you would think that a depth of field calculator could do exactly this, but also take into account situations where the farthest item is much closer than infinity — and you’d be right! It’s not something you’ll find from the charts in my “sharpest aperture” article, but it is indeed possible within the app “OptimumCS,” also by George Duovos. For instance, in your one-to-twenty-meter example, given a 24mm lens, the app tells me to focus at 1.9 meters (note: not double the distance!) and use an aperture of roughly f/14, to result in a circle of confusion that is equal to 0.027mm in both the foreground and the background. This particular difference — .027 vs .03 for the traditional chart — isn’t a huge improvement, but there is still a small difference :)
In practice, many cases will be better. For example, if my closest foreground is 2 meters away, and my farthest background is 6 meters away (perhaps I’m indoors), I would focus at 3 meters (note: not double the distance!), and set an aperture of f/8.5 (in practice, f/8 or f/9) to achieve a circle of confusion sized 0.016mm in the foreground and background — a sharpness improvement of almost 50% over what you would achieve with a hyperfocal distance calculator calibrated to a .03mm circle of confusion. (I didn’t calculate any of these values manually; I just used the OptimumCS app.)
It’s not just this situation, either. Hyperfocal distance charts are designed to give you a set amount of background blur. 0.03mm, 0.02mm, whatever you want — but they don’t take into account foreground sharpness at all, or how to equalize foreground and background sharpness. If you set a stricter circle of confusion value, such as 0.01, it’s true that you’ll get a sharper background. However, your foreground sharpness will necessarily decrease as you focus farther away and use an aperture with less diffraction.
The trick is that different landscapes will let you capture differing total amounts of sharpness from front to back. In a landscape where most of the scene is in the distance, you may be able to set f/8 and get remarkably low circles of confusion in both the foreground and background — something like 0.01mm. But, in a landscape where the nearest element is right next to your lens, it might actually be true that the best you can do (in a single photo, that is) is a much larger circle of confusion — something like 0.025mm, 0.03mm, or, yes, even larger than that, like 0.04mm.
Ideally, there would be a way to determine exactly what the best achievable circle of confusion is for every landscape. And there is! That’s what the George Duovos apps, as well as my “sharpest aperture” article (and a few other, older resources out there) do. Given that you focus at twice the distance to your nearest subject, they tell you the sharpest aperture to use — the one that minimizes the circle of confusion in both the foreground and the background. That aperture can be calculated via the formula I added above (or found by looking at the resulting apps/tables), and it is different from what a traditional hyperfocal distance chart would tell you. However, it is quite accurate, because it changes depending upon the scene, rather than providing the same amount of background sharpness for every image.
And, while you are right that this sort of thing doesn’t matter if it isn’t even perceptible at your print size, I would argue that there’s no reason to avoid doing something that will improve the image quality of the original photo itself. If you later end up printing larger, you’ll be glad that your method was optimal, rather than just good enough to avoid noticing its flaws in a smaller print. That’s all I mean by insisting that there are sharpness compromises every time you use a traditional hyperfocal distance chart (unless the scene you’re in happens to also have a best-possible-case circle of confusion of 0.03mm, which of course can be true, though is no more likely than any other number).
Thank you for your understanding. I hope this makes sense.
Spencer,
No big deal. Although a bit surprised, I wasn’t offended.
Trying to argue a point without sounding negative, is quite a challenge.
I look forward to reading more of your articles.
Erik
I’ve read this entire article and all the comments. Some of the formulae went over my head.
Being a pragmatic guy I went out and tested this in the real world, then pixel peeped to check “sharpness”. All I can say is that even by estimating the double the distance figure, I got sharper results.
I don’t need a spreadsheet to convince me. My own results convince me. If people tried things instead of arguing back and forth they might like or dislike the results – but at least they would know for themselves for sure!
Very glad I read this article and even happier I tested it out.
Happy to hear it, Amin. Good luck with this technique!
I think this article would benefit greatly from having comparison shots between shots focused at the hyperfocal distance and shots focused properly.
Good point! I don’t have any side-by-side comparisons yet. I’ll see if I can make one and add it to the article.
I have benefitted much from the above article and subsequent conversation/comments. Many should heed the respectful manner in which Spencer and those individuals debating the finer points came to close their efforts to ultimately educate. Thank you.
Thank you, Geno, much appreciated.
Hi Spencer. First, I’d like to thank you for giving us another method of focusing foreground to infinity and one that promises to be easier to do out in the field. I had 2 things I wanted to point out to you. After reading ~90% of the comments, I read enough to catch someone else already pointed out 1 of the 2 things I wanted to do: “…you are entirely correct that this double-the-distance method is not accurate when your far limit is something closer than infinity.” Good to see you agree with this, because it had me confused. As far as I’m aware, the benefit of HFD charts is that it makes sure you reach infinity on the far limit for the focal distance and aperture of choice. I see the HFD chart as a cheat sheet, showing me where the near limit is from infinity for each focal length and aperture of interest, leaving me to only worry about where to put the subject or my camera (in order to capture a foreground subject within the focus field). I’ll have to test your method out to see if your best possible aperture charts provide better results vs using the sharpest aperture for your particular lens and a HFD chart approach. I suggest you merge your two articles into one, because this one feels a bit incomplete without the other.
Currently, I use the sharpest aperture for the lens I’m using (thinking that would give me the sharpest foreground to infinity possible). If that would not give me a close enough near limit to capture my subject or a foreground element, based on a HFD chart, only then do I stop down the aperture or move the camera back, but never above the limit where diffraction is visible for said lens. Since this aperture varies between lenses, I’ve always done test shots to find it (lens reviews are usually good in finding it).
The other thing I wanted to point out is that, the DoF calculator I’ve been using accounts for cropped sensor cameras, which is what I use, by using a CoC of 0.020. I always assumed this is equivalent to a FF with a CoC of 0.030, but don’t know if it is. Are your aperture selection charts good for cropped sensor cameras or FF only?
Thank you for your comments! That is true, these two articles are quite closely related. From Google’s perspective, I probably shouldn’t combine them at this point, but I’ll be sure that any future articles on this topic link back to them together, rather than separately.
I definitely do recommend testing out the sharpest aperture charts with the double-the-distance method. I think you’ll find the results to be quite good (in theory, they’re the best that can be achieved, so long as you set up the chart properly). They work equally well regardless of your sensor size, and you don’t have to do any equivalence calculations at all to use them optimally on any sensor. That, alone, is one major reason to use this method :)
It’s true that, generally, a smaller CoC is considered better with smaller sensors, since an equal-sized print will magnify that CoC more compared to larger sensors. However, if you’re specifically making smaller prints due to a crop sensor, 0.03mm isn’t considered bad. Although whether or not you see any blur is down to a number of other factors, including viewing distance. That’s why I’d rather just capture the sharpest photo in the first place and be done with it, rather than calculating whether my sub-optimal aperture and focusing distance are technically “not blurry enough” to notice blur in a print. But everyone has different methods, and personal preferences definitely play a role here.
You beat to it :) I think I get it now…whereas a HFD chart simply tells you the near limit to infinity distances per focal lengths and aperture selections based on an acceptable sharpness, your chart does the same, but based on subject distance and optimizing the aperture for that to infinity (including actual sharpness tests for your lens on the far end)…a better cheat sheet – looking forward to trying it out next weekend. Thanks!
On the short end of the HFD if I don’t want to stack I either put up with noticeable/unacceptable diffraction, switch lenses or simply step back a few feet, whichever works better.
Ignore my last question there…I thought I finished your choosing the aperture article but missed a section…I see that it works regardless of sensor size. Thank you.
You beat to it :) I think I get it now…whereas a HFD chart simply tells you the near limit to infinity distances per focal lengths and aperture selections based on an acceptable sharpness, your chart does the same, but based on subject distance and optimizing the aperture for that to infinity (including actual sharpness tests for your lens on the far end)…a better cheat sheet – looking forward to trying it out next weekend. Thanks!
On the short end of the HFD if I don’t want to stack I either put up with noticeable/unacceptable diffraction, switch lenses or simply step back a few feet, whichever works better.
Confused, confused… still confused…. DTD seems so simple…but is it that simple…
Dear Spencer Cox, in your article you have simply said – Double-the-Distance all through.
Does it mean that irrespective of any Focal Length, your formula stands correct? What I mean is – can I use a 70mm or 85mm lens where you have used 15mm and still get the desired fully focused output? Or is there a limitation to the lenses Focal Length that I should use.
The article never says anything about at what aperture I should shoot. Here I am not saying anything regarding sharpest aperture. Can I shoot any frame at any aperture, following the formula Double-The-Distance. I know you have mentioned saying – use a small aperture, but still will that hold good for any aperture or else – start from f/8 and keep climbing up till f/32? What if climbing up the aperture ladder my output is rendered under exposed?
In short, does it mean – Hyper focal Distance has absolutely no bearing with respect to Focal Length of the lens and Aperture in use. Or do I still have to refer one of the umpteen numbers of smartphone apps created or carry a printed sheet for field work.
Am I missing something?
Will be grateful if you could clear away my doubts.
Thanks
I hope the following is useful…
Suppose the nearest object in the scene is 5 m from the camera, and the furthest object is at infinity or very far away compared to 5 m. We want the nearest object and the furthest object to be equally sharp, in other words, because we can’t get them both razor sharp we want them to be equally blurry. Equally blurry means we want them to have the same circle of confusion (CoC). The only way to ensure this will happen is to focus at double the distance of the nearest object, 10 m in this example. This focus distance is our required hyperfocal distance H.
What we don’t yet know is the actual size of the CoC because it depends on the lens focal length and our chosen f-number N. The formula for the hyperfocal distance is:
H = f^2 / (N × CoC) + f, therefore
CoC = f^2 / N / (H − f /1000) μm, where
f is in mm; H is in metres; and CoC is in micrometres.
Let’s try it with a 20 mm lens at f-stop N = 8, H = 10 m:
CoC = 20^2 / 8 / 9.98 ≈ 5 μm
Now let’s try it with a 100 mm lens at f-stop N = 16, H = 10 m:
CoC = 100^2 / 16 / 9.9 ≈ 63 μm
The CoC is about double the commonly used value of 30 μm. Setting the lens to f/32 will halve the CoC, but diffraction will rob sharpness from the whole scene because the diameter of the diffraction disc (Airy disc) is ~43 μm, which is larger than the CoC at f/32. Hence Spencer linked to his article entitled How to Choose the Sharpest Aperture. It enables us to optimize the trade-off between the blur caused by the CoC and the blur caused by diffraction. The optimum f-stop to use for this example is f/27.
NB: Throughout this comment, I have adhered to the optical physics standard of specifying object space distances from the front principal point/plane of the lens, not from the sensor.
Unfortunately, the writer is confusing depth of field with hyperfocal distance.
Hyperfocal distance is NOT about equal relative sharpness around a focused point as stated above. There is only one definition of hyperfocal distance and it applies to all focal lengths and apertures: it is the distance a lens needs to be focused on to MAXIMISE depth of field for a given aperture so as to still give a high degree of sharpness AT INFINITY. It works on the fact that a lens can not focus a sharp image beyond ininity, so if a lens is focused on infinity there is some “wasted” DoF before infinity is reached, using standard DoF calculations. So the question becomes: what is that nearer distance that the lens can be focused on and still permit reasonable sharpness at infinity? This distance is the hyperfocal distance.
It finds favour in so-called pan-focus lenses which are in fact fixed focus lenses and, in particular, short focal length lenses used in standard 8mm cine cameras whose w/a angle lenses of 6.5mm or 5.5mm have extreme DoF making focusing largely unnecessary. These lenses are, generally, set at the hyperfocal distance for their widest aperture/focal length. Some of the higher quality lenses do permit focusing for more critical work, however.
Quote “the correct hyperfocal distance — is simple. You have to locate the closest object in your frame, estimate its distance from your camera sensor’s plane, and then double that distance.” No. This is not hyperfocal distance. This is simply applying the DoF principle which basically says an image will be acceptably sharp for 1/3rd in front and 2/3rds behind the focused point. In measurable terms, of course, the overall depth of field is dependent upon the chosen aperture. and focus point for any given focal length. In practice, the overall field of acceptable sharpness (DoF) will extend from half the hyperfocal distance to infinity.
@C J Nappoly.
Quote “In short, does it mean – Hyper focal Distance has absolutely no bearing with respect to Focal Length of the lens and Aperture in use.” In fact it has everything to do with focal length and aperture. If you have access to a manual focus film camera lens, or two, you can readily demonstrate the application of this principle, providing the lens has a DoF scale. Simply set the lens at infinity, then note the distances indicated at the aperture marks. Note the distance against f8, for example, and set the lens to this distance. Set against the apertures on the other side of the scale you will see that the infinity setting is now set opposite f8. But the near focus point is now closer than with the lens at infinity. So you “gain” additional depth of field. Do this with a lens of another focal length and you will see that the hyperfocal distance will be different. Importantly, though, the principle is the same whatever focal length lens or aperture is set, and that is simply to maximise DoF for a given combination. I hope this helps.
“This is simply applying the DoF principle which basically says an image will be acceptably sharp for 1/3rd in front and 2/3rds behind the focused point.”
Which you contradict with: “In practice, the overall field of acceptable sharpness (DoF) will extend from half the hyperfocal distance to infinity.”
See my comment number 23.
Hi Terry, thanks for the comments. Some parts of what you’re saying don’t make sense to me, though — that “an image will be acceptably sharp for 1/3rd in front and 2/3rds behind the focused point.” If you’ve ever taken portrait photos at wider apertures, you know that isn’t true, and it’s equally untrue in landscape photography. If that’s what you interpreted my method to be, then I would be fully on your side in agreeing that it is incorrect :)
I gave an example in this article where everything in the landscape was at infinity, yet a hyperfocal distance chart tells you to focus at eight feet. In a case like that — among many other examples — your statement that focusing at the chart’s recommended focusing distance will “maximize depth of field for a given aperture” is clearly untrue. Instead, the focusing distance that maximizes depth of field in this case is infinity — specifically, double the distance to the nearest object, which is at infinity (functionally).
If your goal is to capture the greatest possible depth of field, where the foreground and background are equally sharp, the point at double the distance to your nearest object is always the best place to focus — yes, regardless of aperture or focal length. You can test this by shooting a landscape at something like f/1.8, and trying to figure out which focusing distance will result in an equally blurred foreground and background. I certainly wouldn’t recommend shooting most landscapes at f/1.8, but the principle is identical at more typical apertures. Doubling the distance is almost always the preferred technique, since you’ll typically want to maximize background and foreground sharpness, rather than simply taking every photo at the same focusing distance, prioritizing one over the other.
As for the technical definition of hyperfocal distance, it is exactly what I said it is in the article (same as the dictionary definition): “the closest point to your camera that you can focus, while still ending up with an acceptably sharp background.” But, as I went on to mention, “acceptably sharp” inherently allows for you to decide, yourself, what counts as acceptable. For me and many other photographers, an “acceptably sharp” background is one where the circle of confusion equals that of the foreground. So, as you can see, I’m not altering or denying the definition of hyperfocal distance; I’m simply filling in the meaning of “acceptably sharp” to depend upon the best possible sharpness you can get in a given landscape (since not all scenes have the same maximum amount of sharpness). In other words, a photo isn’t acceptably sharp until both its foreground and background are as sharp as possible. (Your personal standard for “acceptably sharp” might not be the same as mine — i.e., it might not change even as the landscape does — but that doesn’t mean that your definition of hyperfocal distance is any different from mine.)
Lastly, it’s worth mentioning that this article pairs with this one on choosing the sharpest aperture: https://photographylife.com/how-to-choose-the-sharpest-aperture. You need both — the proper focusing distance and the best possible aperture — if you want to minimize the circles of confusion in both the foreground and the background and get the sharpest photo from front to back.
@Pete:
You have explained what is CoC mathematically. The formula for the CoC is directly dependent on f(focal length), N(Aperture) & H(Focus distance). Simple to understand for a Physics major.
In layman’s language let me define the following parameters: (correct me if I am wrong):
CoC (Circle of Confusion): In simple terms, what I gather, it’s the measurement of the blur diameter.
Dof (Depth of Field): Area in focus. Which could also mean – area where CoC is minimum.
Hfd (Hyper Focal Distance): Minimum focal distance which would achieve maximum Dof. Maximum area in focus where CoC is Minimum.
@Terry:
Your definition of Hfd : “the distance a lens needs to be focused on to MAXIMISE depth of field for a given aperture so as to still give a high degree of sharpness AT INFINITY”. I perfectly agree.
This definition factors in lens focal length, aperture and also distance to focus.
So my understanding of Hfd seems to hold good. Hfd is a factor which depends on three variables – lens focal length, aperture & distance to focus.
@Spencer Cox:
My query is still unanswered, if DTD is the way to go, irrespective of focal length of the lens, aperture & focusing distance?
I repeat : does it mean – Hyper focal Distance has absolutely no bearing with respect to Focal Length of the lens and Aperture in use?
As stated in your article above:
Actually, the optimal method is remarkably simple: Find the closest element in your photo. Estimate how far away it is. Double that distance, and focus there. (That’s the real hyperfocal distance, as defined by equal foreground and background sharpness.)
As per above stated:
Scenario: With any given lens (it could be 11mm to say 70mm), at any aperture (say from 1.4 to 32) and let the closest element in the photo be at any distance from 1mm to say 1000ft – if I am to double the distance & focus from 2mm to 2000ft, with any combination of lens & aperture, would that complete range from 2mm to 2000ft constitute the Hfd as stated above in the article & reproduced above?
Further would one get the same output in all the permutation & combinations (lens focal length & aperture) in this said scenario?
Even as per Pete’s equation H = f^2 / (N × CoC) + f, Hfd is dependent on 3 variables,– Focal length of lens, Focusing distance & CoC.
So will it still be okay to say DTD is the real Hfd?
Which tantamount to say that Hfd has only one variable ie – Distance, which directly contradicts Pete’s above Formula.
Thanks
I would like to add the following to my above comment:
@Spencer Cox: I chose the above scenario example, as you have made a universal statement – “regardless of aperture or focal length”.
Thanks for your approach in asking these questions; the way you’re asking them makes a lot of sense to me.
So that we’re on the same page, this is my definition of hyperfocal distance. I assume it’s the same as yours, since it’s also the Wikipedia/dictionary definition: “Hyperfocal distance is the closest point to your camera that you can focus, while still resulting in an acceptably sharp background.”
Everything I’ve written in this article is about the definition of “acceptably sharp,” not about the definition of hyperfocal distance. Personally, for “acceptably sharp,” my standard is having a circle of confusion equal to that of the foreground. And I am entirely free to use that standard, because the definition of hyperfocal distance builds in full liberty to define “acceptably sharp” to your own needs. (I’ll get to the benefits of using this standard in a moment.) Other people choose to define “acceptably sharp” as a 0.03mm background circle of confusion for every photo. Still others define it based upon output size, with the strictness changing as their prints get larger or smaller. No matter what your standard is, we’re not disagreeing about hyperfocal distance — we are arguing the merits of each of these definitions of “acceptably sharp.”
For that reason, I have no disagreements with Pete’s equation whatsoever. If your definition of “acceptably sharp” is a circle of confusion with, for example, a 0.025mm diameter, it’s entirely correct. That’s the equation that will tell you the hyperfocal distance if you have specified a particular circle of confusion in the background, whatever it may be.
Since my standard for “acceptably sharp” is a bit different, it’s true that this equation no longer applies. Instead, I use the equation Hyperfocal distance = (Distance to near object) × 2.
To answer your question, I can categorically say that this equation will always provide you with circles of confusion that are equal between the foreground and background, given that your lens doesn’t have decentering or field curvature issues, and that you aren’t using tilt-shift or bellows features. Some of your examples are quite impossible (an object one millimeter away from your camera sensor would be inside your camera), but I’ll still take a fairly extreme example. Say that the closest element in your landscape is one foot away, so you focus at two feet. Even if you have a 70mm lens set to f/1.4, it is undeniably true that the region at 1 foot will have a circle of confusion equal to that of the region at infinity. So, yes, in this case, and by my definition of “acceptably sharp,” it’s true that the recommended focusing distance would be two feet away. And as crazy as that sounds in this particular example, I stand by it fully. That’s because — in an inexplicable hypothetical scenario where you need the foreground (at one foot) and the background (at infinity) to be equally sharp, yet you’re shooting at 70mm and f/1.4 — it’s totally true that focusing at two feet will provide that.
It’s also entirely true that if you’re using a 14mm lens at f/16, and the closest object in your photo is one foot away, focusing at two feet will once again equalize the circles of confusion between the foreground and background. But in this case, it means more, since those equalized circles of confusion will be quite small, and you’ll actually see detail in those regions :)
So, yes, this method is focal length and aperture agnostic (unless changing to a different focal length alters your composition enough to exclude or include a different “closest object,” at which point the hyperfocal distance is twice as far as your new closest object.)
To take another example scenario, say that you’re photographing a landscape where the nearest object is three meters away. In that case, if your goal is to equalize foreground and background sharpness, you would do so by focusing at six meters. The aperture that results in the smallest total blur in that case is f/8.5 (the calculations for which can be found at the George Duovos link, or the Sharpest Aperture link that I have been posting throughout the comments section). Specifically, the total blur spot diameter in both the foreground and background (since they’re equally sharp as one another) will be .016 millimeters. There isn’t a single other focusing distance and aperture combination that will allow the foreground and background sharpness both to be less than that. And that’s the power of the double-the-distance method. If you prioritize foreground and background sharpness equally, which is what many landscape photographers want in most cases, this is the only focusing distance that renders both as sharp as possible (and, from there, finding the best possible aperture to minimize the total blur — Airy disk combined with circle of confusion — is quite easy).
This distance is by no means the “only real” hyperfocal distance. You’ll notice that it provides different focusing distance suggestions than a 0.03mm hyperfocal distance chart would, which provides different focusing distances from what a 0.02mm hyperfocal distance chart would, and so on. Once again, all of these definitions are equally valid, since they’re the same definition; they simply have different standards for “acceptably sharp.”
But it is true that if your goal is to equalize foreground and background sharpness (and that’s your standard for “acceptably sharp”) then, yes, it is the only real hyperfocal distance.
C J Nappoly,
You wrote:
“CoC (Circle of Confusion): In simple terms, what I gather, it’s the measurement of the blur diameter.”
Yes, it is indeed the blur diameter imparted to the 2D image space of the film/sensor by objects within the 3D object space of the scene which reside further away from, or nearer to, the infinitely-thin plane of focus of the lens. For well-corrected rectilinear lenses designed for general purpose photography, their infinitely-thin object plane of focus lies at the distance from their principal plane to which they have been focussed by the user, or by the autofocus mechanism in the camera). Both the object space plane of focus and the image space plane of focus are, by definition, infinitely-thin on the longitudinal optical axis [the distance/depth axis, rather than the width and height axes]. These planes of focus are infinitely thin, not due to lens focal lengths and/or lens f-stops, it is simply because films and sensors are wholly incapable of recording axial [distance/depth] information — the third dimension of 3D space and of 3D objects [width, height, and depth].
Trigonometry is the branch of mathematics which contains the fundamental tools for mapping 3D object space to 2D image space via a huge variety of projection systems, including the rectilinear projection system commonly used in photography. CoC is one of the many, very useful, derived parameters from rectilinear projection systems. Because CoC is a parameter which is derived from non-trivial mathematics, it can be, and usually is, a pain in the ass for photographers to use it for the purposes of choosing optimal settings for their specific equipment. Nasim, Spencer, and a few others, very kindly distil the complexity of the mathematics down to a form which is not just easier to understand, much more importantly, they present it in a practical form that helps non-expert photographers to improve their images. The technical articles published on Photography Life are not opinion pieces backed by rhetoric, anecdotes, and widely-propagated myths; they are backed by science, mathematics, empirical evidence; and the articles are corrected when errors are discovered — the technical articles are underpinned by the pillars of 21st-century epistemology.
You also wrote:
“Dof (Depth of Field): Area in focus. Which could also mean – area where CoC is minimum. Hfd (Hyper Focal Distance): Minimum focal distance which would achieve maximum Dof. Maximum area in focus where CoC is Minimum.”
Yes, “area in focus” could be taken to mean anything that you want it to mean. Whereas the very specific technical terms “depth of field (DoF)” and “circle of confusion (CoC)” most definitely do *not* mean “area in focus”! DoF relates to axial/longitudinal object-space depth, *not* transverse/lateral area.
I shall reiterate one of my previous statements: Some of the above may appear to be semantic nitpicking, but mathematics and science would be useless if they were not semantically correct, unambiguous, precise, and accurate.
You opened your initial comment on Spencer’s article with:
“Unfortunately, the writer is confusing depth of field with hyperfocal distance.”
I shall likewise close this comment with:
Unfortunately, the commentator C J Nappoly is propagating an endless circle of confusion which is based in only widely-propagated myths.
C J Nappoly, I am always more than willing to provide technical answers to people who genuinely seek answers to their questions. I am also always more than willing to provide satirical replies to muppets.
Apologies for my error. It was the commentator Terry B who wrote: “Unfortunately, the writer is confusing depth of field with hyperfocal distance.”
This article by Spencer for me is like the findings of an astrologer who tries to generalize his findings but fails to cross the limitations of approximation.
Spencer in this article has enumerated his findings and brought forth to us a new dimension for Hfd that is DTD. He is undoubtedly a great photographer. However, from his vast library of pictures that he has shot, he too seems to have penned down his observations , without bothering to validate the established scientific Formula for Hfd in this article. His new hypothesis says DTD is the HFD, forget about all those HFD tables and sheets they are all wrong – The heading of the article says it all.
Come on, Spencer show some mercy for all those scientists & mathematicians who have established the formula for HFD.
Dear spencer innumerous readers must have hugged your DTD approximation without ever knowing anything regarding HFD. Please don’t apply DTD approximations to generalize HFD. I request you to correct your definition of DTD to incorporate dependence of HFD on aperture and focal length of the lens used.
Let me illustrate what I am saying from one of the DTD article written by another photographer like you how it should be done. This is the url where you can find the article: https://jimdoty.com/learn/dof/hyper_chart/hyper_chart.html
Below is the reproduction of the gist:
“For practical use in the field, determine the distance from the closest object that you want to appear sharp to your camera. This is simple. Just focus on the closest object and check the distance scale on your lens. Double that distance to get the hyperfocal distance and focus the lens accordingly. Then use the chart to determine the aperture you need for the lens you want to use.”
This guy also tells you to DTD, but he does not stop there, he further says to check the chart for determining the correct aperture for the focal length of the lens that you are using. This is in conformity with the HFD formula which Pete has penned down.
That is exactly what I have been saying all through.
All the illustrations in this article are shot with lens range from 15mm to a maximum of 24mm, aperture range from minimum of f/8.0 to f/16.
Change the lens to 35mm and aperture to f/5.6, all these illustrations would be busted.
Hence, the DTD hypothesis works only within these following approximations:
Lens range wide to super wide angle lens less than 24mm.
Aperture f/8.0 to f/16, below which diffraction might set in.
Apply DTD only for Landscape photography and within the above approximations.
Most intermediate, advanced and pro photographers need not be told about these so called tricks. This trick or say approximation may be targeted for novice to amateur photographers who may not be aware of these terms.
Hence the above quoted approximation may be included for the benefit of these photographers.
@Pete:
When the going gets tough, it’s time for “Enter the Dragon” & Dear Pete hops in. I have always noticed this from the numerous comments in various other articles of PL, this case being no different. Pete with all his mumbo jumbo terminology tries to scares away the guy commenting. Here too, Rendering 1D, 2D,,,,10D dimensional to the comments were unwarranted, my request, please don’t do this in future. There may be much more well-read audience reading these articles.
I have said what I had to say. A similar article by Spencer on HFD had drawn the ire of many a readers way back. I don’t want to reproduce all that. I am sure you are all aware of that. In that article too Nasim had to intervene to calm down.
It’s up to you to include these approximations to the article listed above or not to.
I rest my case here.
Thanks
C J,
Do you want your background and foreground equally sharp? If you do — and, in most cases, a majority of landscape photographers do — there is no single spot to focus and no aperture to use at all, bar none that will give you sharper results than the techniques I have described. If you need more of an optical science background, and you are unable to understand Pete’s “mumbo jumbo terminology,” read what other photographers have written on the subject — George Duovos, Paul Hansma, Stephen Peterson, and even the other Photography Life article on this subject. (Sources: https://photographylife.com/how-to-choose-the-sharpest-aperture, http://www.largeformatphotography.info/fstop.html, and http://www.georgedouvos.com/douvos/OptimumCS-Pro_Optical_Science.html).
None of this is revolutionary stuff. People have known about it for a long time. This article only became controversial because a small handful of late-arriving readers are doubling down on the techniques they have heard in the past, which are not totally wrong, but also are not perfectly optimal. I think the vast, vast majority of people who read these articles will understand where the facts lie.
Let me offer a challenge. Say that the closest object in your scene is 25 feet away, and you’re using a 50mm lens. Where would you focus, and what aperture would you use in order to result in the foreground and background both being as sharp as possible? And, crucially, please tell me the total blur projected by both the foreground and background regions onto your sensor at your chosen aperture and focusing distance.
I’ll answer the challenge myself to show you how quick and easy it is. (My hope is that your next reply has something similar, since it only takes a minute or two to calculate these values for yourself, and there is no excuse for not including them but to demonstrate that you do not understand the question in the first place.) I would focus at 50 feet and use an aperture of f/11. The total blur in both my foreground and background will be 0.022mm in diameter, a number that accounts for diffraction (which most hyperfocal distance charts do not).
If you know of a way to achieve something that’s sharper than this in both the foreground and background, let me know. This doesn’t have to do with print size, either, because — no matter how large you print — you will never do better than 0.022mm of foreground and background blur in a single image. The sharpest possible settings for a 24×36 print are always the sharpest possible settings for an 8×12 print, too, although it is obviously true that you’ll see errors more easily at a larger output.
If you don’t have your own aperture and focusing distance to posit, there is nothing you can add to this discussion. If you do have a different focusing distance and aperture that result in a sharper image in both the foreground and background, accounting for both diffraction and out-of-focus blur (Airy disks and circles of confusion, respectively), I will firmly apologize and amend this article entirely (as well as the other articles I have written on hyperfocal distance), because it will mean that I have made a huge mistake. Until then, I will stand firmly by what I have written, as, to the best of my knowledge, it contains no misinformation.
(It is also possible that you simply misunderstood this article and didn’t read the other one I wrote. There is an entirely separate article on Photography Life devoted to choosing the sharpest aperture for your scene, once you’ve already found your hyperfocal distance via the DTD method. You can read it here: https://photographylife.com/how-to-choose-the-sharpest-aperture. And, indeed, the sharpest aperture very strongly depends upon your hyperfocal distance and your focal length. I never said otherwise, and, if I did, please let me know where, and I will correct it immediately.)
C J Nappoly,
You wrote mainly rhetoric of the form which I am not the slightest bit interested in engaging with, because it has nothing whatsoever to do with the science and mathematics of hyperfocal distance.
I shall reply to three of your specific statements:
1. All the illustrations in this article are shot with lens range from 15mm to a maximum of 24mm, aperture range from minimum of f/8.0 to f/16. Change the lens to 35mm and aperture to f/5.6, all these illustrations would be busted.
Generally, when shooting a landscape from a standing position, the nearest object in the scene — that is captured within the angle of view of the lens/camera combination — becomes further away as the focal length of the lens is increased. E.g., when shooting a landscape using a 500 mm lens on an FX format camera, the nearest object captured in the scene is very unlikely to be only a few metres away from the camera.
The vast majority of my landscape photography was captured using [either directly, or the equivalent of] lenses between 50 mm and 300 mm on FX format. Many times this entailed finding a vantage point carefully-selected to exclude distracting foreground clutter.
I can think of various reasons why someone would suggest shooting a landscape using a 35 mm lens set to f/5.6 — without specifying the details, and without providing an example image worthy of analysis and discussion! The most compelling reason that springs to mind is: they have inspected MTF charts and discovered that their 35 mm lens seems to be sharpest at around f/5.6. Yes, their discovery would be valid if, and only if, they are using said lens to photograph 2D test charts; and their discovery would be hopelessly invalid if they set said lens to f/5.6 then attempted to use it to capture a variety of real-world 3D object space scenes.
2. This article by Spencer for me is like the findings of an astrologer who tries to generalize his findings but fails to cross the limitations of approximation. … It’s up to you to include these approximations to the article listed above or not to.
The mathematical approximations used in the derivation of the formula for the sharpest aperture are provided in Spencer’s references: the onus is upon you to read the references; the onus is not upon Spencer to provide a plethora of quotations from the references.
3. I rest my case here.
You haven’t made a case, therefore you have no case to rest.
As I stated previously: I am always more than willing to provide technical answers to people who genuinely seek answers to their questions. I am also always more than willing to provide satirical replies to muppets.
I’m guessing you’ve misunderstood that the two comments relate to two different arguments: DoF and Hyperfocal distance. Both are are inter-related but true hyperfocal distance is the maximum DoF for any given lens/aperture combination that provides acceptable sharpness at infinity. With this setting the near point of acceptable sharpness, assuming we use full aperture to start with, will gradually extend closer as the lens is stopped down, but again, always providing acceptable sharpness at infinity. The half distance I referred to is relating to hyperfocal distance.
As Spencer alludes to in his post, what is acceptable sharpness is not written in stone and is subject to a number of variables: actual lens quality, how large is the print, and the distance a print is to be viewed at, and the size of the Airey disc. And not to discount how good our own eyesight is!
This helps to explain we two people can have different views on what is sharp. But to get back to the points I made about 1/3rd in front/2/3rds behind, and half the focused distance when discussing hyperfocal distance. Without resorting to the algebraic equations, I’m getting on a bit too much for these nowadays :D) the comments can easily be demonstrated by using a normal 35mm camera lens that has a DoF scale, or by reference to printed DoF scales.
Taking DoF first, I’m using my Olympus f1.2/55 for this example and which, whilst it has a cramped DoF scale, is still suitable for demonstration purposes. Bear in mind on a lens these actual distances are approximate but good enough for this exercise. I focused on 8′ and at which point the near distance of acceptable sharpness at f16 is 6′ and the far distance is 12′ for a total DoF of 6′, with 1/3rd in front and 2/3rds behind. This ratio of roughly 1/3rd in front to 2/3rds behind holds good for any lens/aperture/focusing distance. Incidentally, I used f16 as this directly lined up with the distance scale in feet. The meter distances didn’t line up as they are displaced relative to the feet settings.
Now we do the same for the hyperfocal distance. Focused at infinity, the near distance is indicated at 6m at f16. Now moving this setting to the lens focusing indicator and reading off the DoF at f16 I get 3m to infinity. That is half the hyperfocal distance. And this simple demonstration demonstrates another immutable law of optics. Whatever is the hyperfocal distance, the zone of acceptable sharpness will extend from half this distance right up to infinity.
Hello, Spencer, Thanks for you reply.
“your (my) statement that focusing at the chart’s recommended focusing distance will “maximize depth of field for a given aperture” is clearly untrue.
I’m certain your misunderstanding of what I said, does emanate from a misunderstanding of what hyperfocal distance actually is. I’ve got a feeling that this could be from a incorrect definition that you have inadvertently latched on to. (See later) What I was saying was actually in regard to hyperfocal distance where at each aperture, setting the the lens at its hyperfocal distance FOR THAT APERTURE, will maximise DoF. I’ve highlighted “for that aperture” as I may not have been abundantly clear in the first instance.
The roughly 1/3rd in front 2/3rds behind argument relates to Dof and holds good at any focal length/aperture/focusing point combination. I have to say roughly because I agree with your comments on why acceptable sharpness can differ. And, yes, if you’d care to do the calculations, it holds true for your portrait example, but you will be working to a much narrower DoF. But without going into the realms of optical formulae, I’m getting on a bit, and even basic algebra can put me to sleep, :D) the following examples may suffice.
Using my f1.2/55 Olympus lens and its DoF scale it shows that when, for example, it is focused at 8′, the zone of acceptable sharpness extends from 6′ to 12′ at f16. (Let’s ignore diffraction for this argument.) So we have 1/3rd in front of the set focus distance and 2/3rds behind of the overall DoF. Testing with a 135mm tele lens I get very much the same ratio. Clearly the lens DoF scale is only a guide and, obviously, published DoF tables will be more accurate but the ratio of in front of and behind still applies. This is a law of optics.
Now if I do the hyperfocal test with the same lens, I get a hyperfocal distance of 6m using f16. Now when I set the lens at 6m the DoF scale indicates a zone of acceptable sharpness extending from 3m to inifinity. This is the basis of my earlier post, i.e, the zone of acceptable sharpness runs from half the actual focused distance to infinity.
@ Pete A. Does this help your understanding of what I said?
@ C J Nappoly. Yes, it is that simple. I’m not sure why it is causing so much confusion.
@Spencer.
“Actually, the optimal method is remarkably simple: Find the closest element in your photo. Estimate how far away it is. Double that distance, and focus there. (That’s the real hyperfocal distance, as defined by equal foreground and background sharpness.)”
This is NOT the hyperfocal distance, which requires a calculation for infinity, nor is it an application of DoF. This as a definition is very woolly. No mention of aperture nor focal length. For example, using the above, my guesstimated, or even measured distance is, say 5′. Double this to 10′ and focus there. Now how does this tell me what the DoF is? It is not a measure of hyperfocal distance, as it stands, it may be by coincidence, but without knowing focal length and aperture it tells us absolutely nothing other than whatever subject is at 10′ will be sharp.
I’d be interested in knowing whence this definition came.
Regards,
Hi Terry,
Let’s assume that you are comfortable with a 0.03mm circle of confusion in the background. Taking that standard, I’ll also assume you’re using a focal length of 24mm and an aperture of f/16. A hyperfocal distance chart (calibrated to 0.03mm) will then tell you to focus at 4.02 ft, with the region at 2.01 feet also having a circle of confusion of 0.03mm.
Since the distance behind the point of focus that is acceptably sharp is infinite (infinity minus 4.02 feet), and the acceptably sharp distance in front of the point of focus is 2.01 feet (4.02 minus 2.01), I think I still don’t understand your quote: “an image will be acceptably sharp for 1/3rd in front and 2/3rds behind the focused point.”
I think once we clear that up, it will be easier to figure out what each of us is saying.
Also, that isn’t my definition of hyperfocal distance — that’s my method of finding the hyperfocal distance. My definition is simple, and I don’t think you disagree with it (if you do, please let me know): “Hyperfocal distance is the closest point to your camera that you can focus, while still resulting in an acceptably sharp background.”
I’m curious what your personal threshold for “acceptably sharp” is. Is it related to output size? Would you be comfortable saying that, at a huge output size, the acceptable blur would have to shrink to a stricter standard? And if you do agree with that, are you willing to accept the almost inevitable sharpness tradeoff in the foreground that comes from using a stricter standard in the background, since you’d end up focusing farther away and using a wider aperture (to minimize diffraction)?
For now, I think we’re still out of sync on some basic definitions, and we should straighten that up before going into a more technical discussion.
Terry B,
You wrote: “The roughly 1/3rd in front 2/3rds behind argument relates to Dof and holds good at any focal length/aperture/focusing point combination.”
I have previously explained this widely-propagated myth in my comment number 23, in which I busted this myth using both science- and empirical-based evidence.
You wrote: “Using my f1.2/55 Olympus lens and its DoF scale it shows that when, for example, it is focused at 8′, the zone of acceptable sharpness extends from 6′ to 12′ at f16. (Let’s ignore diffraction for this argument.) So we have 1/3rd in front of the set focus distance and 2/3rds behind of the overall DoF. Testing with a 135mm tele lens I get very much the same ratio. Clearly the lens DoF scale is only a guide and, obviously, published DoF tables will be more accurate but the ratio of in front of and behind still applies. This is a law of optics.”
Of bleedin’ course: the harmonic mean value of 6 and 12 is, by definition, precisely 8 — irrespective of the units of measurement.
Using the details provided in your comment: a 55 mm lens used at f/16 will provide ‘a range of acceptable sharpness’ around a focus distance of 8′, from an object space range of distances from 6′ to 12′, when using a CoC of 0.026 mm. The hyperfocal distance of the lens under these conditions is precisely 24′; the focus distance to which the lens has been focussed, 8′, is precisely one-third of its hyperfocal distance. Thank you for confirming my comment number 23!
Set your 55 mm lens to a focus distance well away from 8′ at f/16 and its DoF scale will totally bust your claim of “roughly 1/3rd in front 2/3rds behind”.
Why are you propagating this thoroughly-debunked myth on a website which is dedicated to providing science- and evidence-based information [rhetorical question].
Hi, Spencer. I think we are at risk of befuddling each other. :D) We don’t need to play around with formulae, it’s quite easy, simply with a lens in hand, as you have demonstrated with the 24mm lens. I do have a 24mm Elmarit-R but I’ll take it as read what your figures indicate.
Your quoted hyperfocal definition didn’t pin down what it truly represents. It’s 95% there, but the missing 5% would fill in an important gap. Let’s look at it again. “Hyperfocal distance is the closest point to your camera that you can focus, while still resulting in an acceptably sharp background.” Background? Where is this plane situated?
I’ve found a hyperfocal explanation in one of my Leica books, but it is simpler to grab from the internet, and interestingly, I found this from wiki:
Definition 1: The hyperfocal distance is the closest distance at which a lens can be focused while keeping objects at infinity acceptably sharp. When the lens is focused at this distance, all objects at distances from half of the hyperfocal distance out to infinity will be acceptably sharp.
Definition 2: The hyperfocal distance is the distance beyond which all objects are acceptably sharp, for a lens focused at infinity.
The first of these is the one that has generally been put forward over many decades, and is the one I have been basing all my comments appertaining to hyperfocal distance. I’ve not seen the second definition used in over 55 years pf photography, but I would argue is not strictly correct as it doe not maximise DoF. In practice, it is very close, and not really worth arguing the point as to how the total extent of DoF slightly differs.
So what is missing from your definition is the reference to infinity. “..an acceptably sharp background” is not saying that it is at infinity. What if the subject is a group of people in front of a building 50ft away? The building is the background but definitely not at infinity. If the argument now runs to ensuring acceptable sharpness for the building (or conversely, it isn’t wanted to be sharp) one needs recourse to DoF, easily adjusted using the DoF scale on the lens, or by reference to tables. This is now where my referring to a zone of acceptable sharpness that will have 1/3rd of the total in front of the focused point, and 2/3rds behind it. As you alter the focus point, or the aperture, the extent of overall DoF will, of course, change. But the ratio will always apply. It is a law of optics.
This again, can easily be verified using any focal length lens with a DoF scale to establish the best point of focus for the main subject so as to have the building sharp or put out of focus. If the building is not required to be sharp at all, simply forget DoF, focus on the subject and select the widest aperture. It may even be advantageous to actually move you subject so it is actually on the point of critical focus and control DoF with a suitable aperture. But remember we are now talking about using DoF, NOT hyperfocal distance, and the 1/3rd – 2/3rds relationship still holds.
Quote: “I’m curious what your personal threshold for “acceptably sharp” is. Is it related to output size? Would you be comfortable saying that, at a huge output size, the acceptable blur would have to shrink to a stricter standard? And if you do agree with that, are you willing to accept the almost inevitable sharpness tradeoff in the foreground that comes from using a stricter standard in the background, since you’d end up focusing farther away and using a wider aperture (to minimize diffraction)?” To answer these, I’d reply as follows.
If you’ve read enough photographic books on what is and what isn’t sharp you’d soon realise that there is no one precise standard, and virtually all statements on the matter use “acceptably sharp”. I can only say I know that what appears acceptable to me may not appear so to many others. Irwin Puts is the world’s acknowledged expert on Leica Lenses and he does touch on this subject in his compendium when he refers to the human eyes’ poor resolving power, and which gets less discriminating over distance. Basically, this means that a photographic print needs to be at its sharpest when up close, but we lose our ability to assess sharpness the further the print is away from us. It follows, that a large print on a wall several feet away can appear sharp even though were we to examine it close up, we’d definitely see that it wasn’t. What is acceptably sharp has everything to do with viewing distance.
Look at those huge advertising posters you see around town. Looked at from a distance they look wonderfully sharp. Get close, though, and the actual resolution is appalling. This is a good example of what Puts (and other experts) is saying. In other words, beyond a certain viewing distance, and it is fairly close, the human eye has difficulty distinguishing any increase in sharpness (resolution). In my poster example, it would have to be at a much higher resolution when viewed close up, but at this new resolution, it wouldn’t make it appear any sharper if, say, viewed at 25ft, as the eye cannot make out that resolution at that distance.
I see I’ve been going on somewhat, but I trust you’ve made it to here. :D) As a punishment, I’ve now got another tortuous captcha session!
Fair points! You are right, I do have an error in the definition of hyperfocal distance that I gave you; in trying to make it more colloquial, I substituted “infinity” for “background,” which indeed is not always true. So, a revised definition (hopefully we now fully agree!) is “Hyperfocal distance is the closest point to your camera that you can focus, while still resulting in the region at infinity appearing acceptably sharp.” I’ve now corrected this definition in the original article as well; thanks for that.
You’re also quite correct that viewing distance is a major factor in whether or not an image’s final output appears acceptably sharp. Same with output size, personal ability to distinguish detail, output material, and so on. Quantifying all of these variables in a meaningful way wouldn’t be very practical, although there certainly have been attempts to do so.
However, your statement on the 1/3, 2/3 zone of acceptable sharpness is still mistaken. To quote the same Wikipedia article, “The hyperfocal distance has a curious property: while a lens focused at H [hyperfocal distance] will hold a depth of field from H/2 to infinity, if the lens is focused to H/2, the depth of field will extend from H/3 to H; if the lens is then focused to H/3, the depth of field will extend from H/4 to H/2.”
Putting that into an example, say that the hyperfocal distance in a landscape is 12 feet. When focused at 12 feet, the region from 6 feet to infinity will be acceptably sharp. However, if you then focus at H/2 (six feet), the region from H/3 (four feet) to H (twelve feet) will be acceptably sharp — which is equal to two feet in front of the focusing point and six feet behind it, a ratio of 1:3. And if you focus at H/3 (four feet), you’ll end up with the region from three feet to six feet appearing sharp — which is equal to one foot in front of the focusing point and two feet behind it, a ratio of 1:2. Focusing even closer, this ratio diminishes further and further. It is by no means a constant 1/3-before and 2/3-after relationship. The relationship varies greatly depending upon how far you focus.
“[Terry B]
Definition 2: The hyperfocal distance is the distance beyond which all objects are acceptably sharp, for a lens focused at infinity.
…
I’ve not seen the second definition used in over 55 years pf photography, but I would argue is not strictly correct as it doe not maximise DoF.”
I’m very pleased that you have not seen that definition used in over 55 years of photography because it was abject nonsense then, and it still is. However, in the scientific field of ophthalmology, the concept of ‘depth of field beyond infinity’ was, and still is, used in both eye tests and in the prescriptions for spectacles: of course, ophthalmologists use the correct terminology, rather than the hilarious terms used by amateur photographers.
I have an inkling that you will continue to argue ad infinitum :-)
Excellent article. Very helpful. Thank you.
A suggestion: Consider sharing those two tables as spreadsheets in Google Sheets or another similar sharing platform. It could save interested readers some time (I just finished re-typing them).
Either way, thanks again!
Glad you enjoyed it! I believe a couple readers in that comments section created their own spreadsheets that can be downloaded. Fair point, though — I know your pain, because it took me quite a while to calculate all those values and type up both the metric and imperial charts :)
@ Spencer.
I think we are there. :D) It just befalls me to qualify the clear misunderstanding I created in my DoF comments and the 1/3rd, 2/3rds relationship. What I do believe, but overlooked to make the point, is that this relationship should hold but only up to where the zone of acceptable sharpness at its furthest point falls short of infinity.
@ Pete A. Does this hold true, as I believe it does, or does the ratio break down earlier? Ad finitum is better than ad nauseum. :D)
Terry B,
Use a DoF calculator, such as DOFMaster, for the purposes of both confirming and denying your beliefs. NB: Confirming our beliefs is a trivial exercise because we can always find plenty of specific examples which support our beliefs. However, the scientific method mandates that we must perform experiments which are designed to disprove/reject our preconceptions and hypotheses. This core epistemic pillar of science is the very pillar which enables science to be self-correcting.
It seems to me that the major problem with the subject of photography is that its explanations omit the prerequisite fundamental steps of learning about the nature of light and optics. Instead, photography starts from focal lengths, distances, and angles of view. Which is very unfortunate because the learner is misled to falsely believe that lenses ontologically possess these epistemological attributes. Whereas, the only things which lenses can actually do are to refract, diffract, and disperse whatever light happens to be impinging on them. Lenses are distance agnostic: they don’t actually have a focal length, they have only optical power! When the optical power of a lens is used for the purpose of forming an image, we often use the thin lens model approximation, combined with geometry, to map object space parameters to image space parameters and vice versa. In this overly-simplified model, the focal length is the reciprocal of the optical power of the lens. Again unfortunately, the thin lens model itself leads people to develop a plethora of false beliefs, which have become widely-propagated myths. I find this both annoying and deeply saddening because I’m passionate about science, mathematics, and critical thinking skills.
I fully endorse Daniel Dennett’s statement on the subject of ignorance:
“Ignorance is nothing shameful; imposing ignorance is shameful. Most people are not to blame for their own ignorance, but if they willfully pass it on, they are to blame.”
Yes, I fully endorse that statement, therefore, I’m always willing to help people who genuinely wish to correct their mistaken beliefs and their false assumptions.
Does this formula apply to APS sensors?
Yes, it applies regardless of sensor size!
Hello Spencer,
Thank you for the well written article and the very helpful DTD guidance. I also particularly appreciate your patience and respectfulness with which you respond to numerous comments. As a rank amateur who takes photos for the fun of it, I never knew that focusing a landscape scene could involve charts, hyperfocal distance and circles of confusion and that professional photographers go to such lengths to get sharp images as a matter of course.
The DTD technique relies on measuring or estimating distance; the lenses I use typically do not have a distance scale and those that do are not that reliable, which presumably leaves us to guess-timate the distance. Is that right? or is there some other distance measuring technique that I am not aware of? Thanks again for you excellent efforts.
Thank you, glad that you have found it useful!
My lenses don’t have distance scales either. Yes, you’re right, the best method is to estimate. If your foreground looks about five feet away, just do your best to find something in the landscape that is ten feet away. Units don’t matter here, of course — if your foreground is two paces away, focus on something that looks like it’s roughly four paces away. Or, if your foreground looks arm’s length away, imagine two arms lined up, and focus there. (One technique you can use on some occasions is to walk into the landscape and literally count paces. But this is only possible if it doesn’t harm the area that you’re walking.)
Because it relies on estimation, there is some inherent imprecision to this method. You might focus a few inches beyond the ideal spot, for instance, or a few inches in front. Still, it’s the best way to focus if you want your foreground and background to be equally sharp. You can always bring a measuring tape or a laser rangefinder if you’re really intense, but I tend not to recommend it, since estimation works quite well :)
I have been taking photos, mainly landscape photos, for more than 40 years. I have always wanted a reliable, simple and easy way to calculate a focus point to get maximum DOF. I have tried carrying DOF tables into the field with me, I have heard about and tried to use the 1/3 in front, 2/3 behind idea, I have tried always shooting at F/22 and I have tried numerous other ideas. None of them worked very well. When they did I began to believe it was an accident.
I am not an optical engineer, I can’t really explain defraction, and I couldn/t explain circles of confusion to anyone. But I do know this, the information in Spencer’s article works. I have taken hundreds of images since reading the article and my success rate of sharp images has been astronomically better. So my point is, I can’t understand a lot of the discussions on this thread, but I don’t need to. I have the results in my image catalog. Sharp images.
Thanks, Spencer.
Ron
Thank you, Ron, I’m very happy to hear that this method has worked so well for you. That’s always the best way to determine the merits of a new technique — testing it in practice.