After I posted my last article comparing the high ISO performance of the Nikon D4s vs D4, a number of our readers requested that I provide a similar comparison with other cameras such as the Nikon D600/D610, D800 and Df. Instead of posting multiple articles that show these comparisons, I decided to put it all into a single article, so that our readers could look at the side by side comparisons, or download the files to their computers for closer examination. Before you start comparing the below images, however, I would like to point out that the images are just provided as a reference, and only represent one side of the camera performance – high ISO in low-light, indoor conditions. Each camera comes with its own set of features, strengths and weaknesses, so please do not draw conclusions from these shots. Please note that ISO performance might vary in different lighting conditions. My recommendation would be to read the comment exchange I had with Brad Hill of Natural Art Images in the previous Nikon D4s vs D4 comparison article, where we discuss the topic of comparing sensor performance in detail.
For those that want to find out how the below images were taken, here is some technical information:
- All images were shot in 14-bit RAW and converted using Adobe Lightroom. Nikon D4s images were processed using the Adobe Camera RAW 8.4 RC
- Camera settings were identical between all cameras, shot in full manual mode (same aperture, shutter speed and ISO). White balance was also set to the same value across all cameras
- Images from the D600 and D800E were down-sampled to 16 MP to match the resolution of the Df/D4/D4s
Table of Contents
1) ISO 1600 Comparison
Let’s start out by comparing ISO 1600 between all cameras:
As expected, all cameras handle ISO 1600 extremely well. The Nikon D600/D610 and the D800/D800E have a resolution advantage here. Images appear sharper, thanks to the down-sampling process.
2) ISO 3200 Comparison
As we push ISO to 3200, noise obviously becomes much more apparent, especially in the shadows. As expected, the Nikon D4s looks the cleanest in the group.
3) ISO 6400 Comparison
ISO 6400 is pretty noisy for all cameras, but the high resolution cameras cannot maintain the same dynamic range as the lower resolution cameras. This is visible when looking at the shadow areas. The Nikon D4s once again looks a tad cleaner, followed by the Df and the D4.
4) ISO 12800 Comparison
ISO 12800 is already beyond the native ISO sensitivity of the D600/D610 and the D800/D800E, so we see big losses of colors and dynamic range there. The D600/D610 looks the noisiest to me with some visible artifacts, but seems to retain colors a tad better than the D800/D800E. The output from the D4s looks very similar to the Df.
5) ISO 25600 Comparison
The D600/D610 looks the worst at ISO 25600 among the group, although the D800/D800E is not that far off. Lots of artifacts, loss of colors and detail. The Nikon Df shows some artifacts in the shadows, but the image is much cleaner in comparison. Once again, the Df and the D4s seem to be pretty similar in performance.
6) ISO 51200 Comparison
ISO 51200 is already beyond the limit of the D600/D610 and the D800/D800E, so those are excluded. While all three are similar, the D4s seems to have a slight advantage over the D4 and the Df.
7) ISO 102400 Comparison
102400 is way beyond my “acceptable” range, so I do not particularly care for the performance here.
8) ISO 204800 Comparison
And ISO 204,800 is just a marketing gimmick, since most of the details are missing, colors are gone and noise patterns are all mixed up.
Summary
Based on what I see from the above, it is pretty clear that high resolution cameras do struggle with noise at high ISOs above 3200 when compared to lower-resolution sensors. If I were to pick the order of best to worst up to ISO 25600, I would say that the Nikon D4s looks the cleanest to me, followed by the Nikon Df, D4, D800E and the D600. However, noise is not everything – as you can see from the above comparisons, the D600/D610 and the D800/D800E have a resolution advantage over the Df/D4/D4s. When images are resized/down-sampled to 16 MP, those cameras are capable of bringing out greater detail. Similar to what happens when you take a somewhat blurry photo and make it smaller for the web to make it appear sharper, the down-sampling process certainly helps in not only increasing the details, but also effectively reduces noise.
Hey Nasim — Good look at their ISO capabilities I’m guessing that the cameras go in the following order:
D600 — D800
DF — D4
D4s
I may have missed it in the article, if I did excuse me.
Yes, that’s the correct arrangement. Just move the cursor over each image and a label with the name of the image pops up.
Would be interesting to see how the D3s fares against this lineup, in case you still have access to one.
Hi Naseem, how does ISO react in 70D cannon? My pictures appear reddish in indoor light… Pls advise thx
can you do one with Canon 6 d and Canon 5D mark III and canon eos 650d?
I know we don’t do much Canon here but I too would LOVE to see what the Canon bodies are producing in relation to our Nikon bodies.
seconded, would be great to see Nasim!
I’d love to see the comparison as well, especially against the Canon 6D. I just shot a wedding using a 6D, a 5DMIII, and my D800E. I was using a slow lens on the 6D (a 70-300L 4-5.6), which forced me to shoot at ISO12800. I was shocked at how good the shots came out. The noise was very fine and film like. Detail and color retention was excellent. I had faster lenses on the 5d and D800E (f/2.8 and f/1.4, respectively), so I didn’t have to shoot them at shut high ISOs, but I doubt that either lens would have performed as well.
It would be great to see how the 6D compares to these 4 Nikon FX cameras.
yes, that would be wonderful!
Deb, since I personally shoot Nikon, it is easy for me to do these kinds of comparisons, as I own 3 of the camera bodies in this list. With Canon, I will have to request each body separately and it not only takes time, but also sometimes impossible to get so much at once. I am definitely planning to cover more Canon gear in the future though. Thank you for your feedback!
A wonderful and very useful comparison for anybody in the market for these bodies, thank you Nasim! From my perspective, and though I’m not alone I also know there are plenty who fall outside of it, my interest in a body with good noise performance really needs to be how it holds up at about ISO 3200 thru 12,800. Anybody above or below is just not that critical as I would rarely ever need to go above, and below we know the results are going to be what they are on a $2000+ FF camera which is; pretty darn good.
With that said, if we’re not printing, so much distribution is now on the web that output resolution really need only be in the 3-4Mp range. When it comes to printing, we have such wonderful noise reduction software available to us built into Lightroom (and Photoshop to some extent) that I also can’t help but wonder if these differences are really that big considering these two important factors.
Nobody who does these comparison actually shows you “Look, when you reduce noise in Lightroom, this is the image you get…” We all want to know what it looks like “SOC” yet none of us really deliver “SOC”. So what’s the point?
Yes, many of us print (I do too occasionally) and yes, I want the cleanest image I can get. But from what I can tell above, 3200-6400 with some noise reduction help and even just a little bit of down sampling and it would take only pixel peeping photographers to see a difference with a magnifying glass yes??
Just my two cents…
Jon, you got that right! For a $2K body, the D600/D610 ISO performance looks pretty darn good! I guess you could draw a conclusion from seeing these crops – at the end of the day, ISO performance across most Nikon cameras is about the same! So when looking at a camera, you should be evaluating your particular needs in regards to AF performance, features, FPS, etc.
In regards to noise reduction, there are too many ways to “skin the cat”. Lightroom and Photoshop have built-in tools, there is very effective selective noise reduction software out there (Nik Software Dfine, Noise Ninja, etc). The reason why I exclude noise reduction, is because the software can react differently to each image and that creates another potential variable that can affect the results.
And lastly, your last paragraph describes the situation pretty well – all these cameras are perfectly capable of great ISO performance up to ISO 6400, maybe even 12800.
Nasim, agreed… your comparison fundamentally shows how little difference there is in high ISO performance, not how much. And although they weren’t part of this comparison I’d add also that this extends to the D700 or D3s. There’s been almost no appreciable gain in high ISO noise performance since 2007.
Based on this if I had to regularly shoot in low light without flash I’d choose a camera whose AF systems and controls were better suited to low light, not a camera that had a half stop noise advantage at ISO 12800. For those currently in the market for an ideal low light camera this rules out the Df and D610 (for their second-tier AF systems) and the D800 for its inconsistent AF with fast primes. The D4s, D4, D3s, D3, and D700 are all better low light camera choices.
I checked only the ISO1600 and ISO12800 samples and discovered in your pictures that the D600 picture has a more sharp detail than the others. To compare that I espacially looked at the white printed order number from the bottom disc which is 105906.
As they seem to me the sharpest on D600 I would like to ask you if you can eliminate that your pics have focus issues?
On the other hand my personal conclusion is that I would need to jump from my D600 to an D4s if I would need a perk that is clearly visible to me. So thanks to make me happy with the things I already own :-D
Mario, actually, if you look at the 105906 number, the sharpest among all is the Nikon D800E. Due to its massive resolution of 36 MP, when the image is resized to 16 MP, it can bring out the most details. The second sharpest is the Nikon D600. Focusing was precisely the same for all shots – I literally swapped the lens on each body without touching the zoom or focus rings. So there are no focus issues. The reason why Df, D4 and D4s do not look as sharp, is because you are looking at 100% view, while D600 and D800 are made smaller in Lightroom/Photoshop to match the field of view of the Df/D4/D4s.
Not that I’m not grateful for this review, but it would have been cool to see a D700 & D7100 thrown in here for comparison. :) That would give a lot of people (like myself) who have the older generation FF bodies or a current generation crop camera to get an idea on those differences.
Jon, and you can find those comparisons in various reviews on the site. Here is a comparison between the D600 and the D700: photographylife.com/revie…kon-d600/7
A similar comparison of the D700 with the D800 can be found here: photographylife.com/revie…kon-d800/7
Forgot D7100 comparisons: photographylife.com/revie…on-d7100/7
Darnit Nasim, now I’m one of those people who “doesn’t read”. Sorry, you’re right, those comparisons are perfect. Thank you!
No problem my friend!
Have a great week!
So, from my summation between the D800/D600 to the D700 to the D7100 we’re looking at a about a stop down in ISO performance in the 3200-6400 range at each level. A D800 will give roughly two stops better performance. Good to know!
If would be fun to run the ISO 6400 and above shots through DxO, and use their PRIME noise reduction option, just to see what the results look like. For what it is worth, if I had the original NEF files, I could do so.
Just saying …
WEJ
William, I did not want to mix noise reduction into such a comparison, because I would have to equally apply it to all cameras and that would deviate the results. I did not post NEF files, because they are all huge, but if you want, I can give access to those via Copy.com to you. Let me know if you want them – will happily share!
I would like to see what effect it would have on the files, so if you are willing to share the NEFs, please do so. I will process each the exact same way in DxO, first with regular noise reduction, and then second with PRIME noise reduction. I will then post to a hidden page on my website, and provide you the link, so that the pics could be downloaded. As I don’t normally down-size, I will let you do any such process (DxO does not have such an option).
If would be interesting if someone else would do the same process in Photoshop, and run the pics through Topaz (I think that is a noise reduction add-on for Photoshop). A comparison of Topaz against DxO PRIME would be interesting to see.
I would prefer to work with ISO 3200 and above, but am perfectly willing to start with ISO 1600 or less.
Please let me know,
WEJ
I am a DxO Optics Pro user, and I appreciate the PRIME technology, introduced with version 9 (November 2013), although It’s a CPU-devourer. PRIME is tightly integrated with DxO Optics Pro and works in the phase of RAW development.
I am not able to say whether it is the best (the best is always a question concerning specific conditions of use), but I can definitely say that PRIME belongs to the few most valuable denoising tools currently available.
PRIME technology is very useful at medium ISO to obtain large prints “pixel peeper resistant”, when the details, the tones and the colors are not affected by the rising sensitivity.
However, the images that you get at higher ISO are of low quality because of the lower amount of useful information in the file.
This is common to all denoising software. There are no miraculous solutions that can replace the information captured by the photodiodes. It ‘s just possible to exploit them better or worse.
An observation.
Despite being very satisfied, I consider DxO Optics Pro an instrument of niche.
For me it is very useful because I take mainly pictures of architecture or landscapes that include architectural elements. For most photographers I think it is more useful DxO ViewPoint 2 (also released in November 2013) that can be used standalone application or as a plugin in LR, PS or Aperture.
Just curious, but is your policy to address most camera systems, except canon? I see that you discuss Nikon the most, but I’m pretty sure you’ve reviewed/discussed a few other lines, but never canon.
Arnel, not just a policy it seems, but a bias. It’s all Nikon digital all the time here. No Canon. No film. No 35mm. No medium format. No large format. All specs, test charts, pixels, computer analysis, and gear. Par for the course.
It is most disappointing. After all, why does he need so many contributors for mostly talking about Nikon. Most people whom I follow, if they are really good, are capable of speaking about more than just one line, no matter what their preference. And besides that, there is plenty of material to cover in the world of photography without talking about any brand at all. Just my two cents.
Arnel, as much as we want to diversify, it just happens that we mostly attract Nikon shooters here – and that’s why there is more of Nikon content vs Canon. Both are equally very good, so by no means do we ever say here that Nikon is better. In fact, I try to avoid cross-brand comparisons for this reason – there are just too many haters on both sides. And while I do appreciate your feedback on “speaking about more than just one line, no matter their preference”, how could someone write about a brand they do not use, or perhaps never used before? I try to cover Canon in many of my articles on photography, but I stay away from making remarks on something I am not 100% familiar with. Lastly, while we do have plenty of gear articles here, if you explore the above menus and look at various sections on beginner articles and post-processing, you will find plenty of non-gear articles :)
If he included Canon, then he may include Sony, Pentax, Fuji, Olympus … I don’t see canon being any more important than any of those. He uses Nikon, that’s what he tests … it’s pretty simple. There are tonnes of Canon specific sites if you try google.
Bias towards what, Nikon? Yes, this site is centered around Nikon mostly, because that’s what I and a number of other PL team members shoot with. I never said that Nikon is better than Canon though – just a system of choice, that’s all. Actually, Laura Murray in our team does shoot Canon and also shoots a lot of film. You can check out her film work in our Gallery and see some of her articles she posted here before.
I think the reason is mostly that Nasim shoots Nikon everyday, and probably owns more than $30000 worth of Nikkor. He’s an expert on everything Nikon, and is much less knowledgeable on Canon, so doesn’t feel like writing about it.
Eric, thank you for clarifying that. Yes, Nikon is a system of choice for me personally, which is why I mostly talk about Nikon here at PL. The only experience I have with Canon is 5D Mark II and Mark III (which I reviewed a while ago), so I try to stay away from discussing something that I am not 100% familiar with.
I don’t think this website have the funds to buy or rent canon equipments, unless B&H or other store would lend them some. They have invested in Nikons to use in their wedding shoots, etc. It’s either you use Canon or Nikon and not yet the other brands. They started with Nikon that’s why they are stuck it, likewise you are stuck to Canon. If other website can be like a jack of all trade and feathers testing all brands, I don’t think they would really give an accurate report in a particular brand or other brands, unless maybe if they are a big company like DXOmark. this site is now looking at the potential of mirrorless cameras and the likes, not 35mm or medium format, like I say they don”t have funds, no offends Nasim or maybe I don’t really know you have some funds. They made an article on to what went wrong with the Nikon 1 system’s very bad price drop, Nikon is doing it again with the Nikon 1 V3 with a high price and photography life wonders what is Nikon thinking? I’m actually a painter and like to look at websites and I don’t personally knows Nasim, English is my second language.
Bes, you are absolutely correct. Getting a hold of a single Canon camera means that we also have to request lenses and potentially other accessories. There are budget limitations and we often cannot get too much on our hands.
But I do have plans to finish reviewing Nikon lenses, then keep only one body and switch to Canon for future reviews. I realize that Canon also has a lot to offer and our readers I am sure are anxious to see more info on Canon.
Arnel, Nikon is what I personally shoot with, so most of the reviews that we have here at PL are Nikon. We do cover some Fuji, Sony and I have reviewed the Canon 5D Mark III before (you can find it in the camera reviews section), but not a whole lot of Canon for sure. Hopefully, we will be able to do more of Canon gear soon – definitely have that in the plan!
It’s fun to see there’s always people complaining. Nasim is doing his best and he owns Nikon gear mostly.
He can’t review all the camera’s that are out there. We all have busy life so is Nasim. Instead of complaining do a review yourself and share it. You will see all the time it consume to do a ”proper” review.
Let’s be grateful for Nasim works and im sure there is other website who talks more about canon.
Actually, I’ve never seen a blog devoted all to canon equipment, though there must be one. Point is, that’s not what I’m looking for. I can search out articles about canon equipment or people opinions, don’t need a blog for that. I began following because I thought is was about PHOTOGRAPHY, not equipment. My bad. I can unsubscribe, and leave all you people alone.
The camera is part of photography as much as the surf board is to the surfer, the brushes and paints are to a painter, etc. The equipment is necessary and valid. It also lends itself well to online discussion and helping keep competiton strong which keeps producers in check.
There is plenty of great photography tips and information here along with discussions about gear. I would encourage you to stay!
I know how you feel. It seems like the majority of sites I visit, unless I specify Nikon, primarily cover Canon gear. Also, as a dSLR devotee, it kinda gets on my nerves to visit sites that go on and on about mirrorless cameras. I have nothing against them…just don’t care to read about it.
Anyway, you don’t have to leave us alone. A reasonable, dissenting opinion is always welcome :-)
Arnel, go to the website index and you will find a great deal of fantastic articles done on photography. This specific one was done with a spicific topic in mind following a number of requests. 90% plus is about Nikon as the owners primarily use Nikon. The reason many people read and contribute is because they bought/ invested into the Nikon system and not any of the other. There are many that constantly moun and complain about issues but I wonder if they ever go out and make an effort to take pictures, let alone sell any. There is also a reluctance to do field comparisons between brands as it normally start an attack on the validity of the test, that the tester is just bashing a brand, and many more. That is what DXO does and very few Canon users trust the data integrity due to the fact the Nikon and Sony sensors dominate the rankings. So there is never a fair comparison as they are different brands with even different objectives.
Thank you so much for this very useful comparison, Nasim.
I agree with Mario that to my eye the D600 looks the sharpest. As a D800 owner (and a very happy one) I find that looking at my pictures in Lightroom or Photoshop, I prefer not to go above 1600. And frankly, I prefer not to go over 400 if possible. This may be psychological, however – a holdover from film days. I haven’t yet tried printing any images with ISOs above 400. Will do so in the future, however.
Thank you again. And thanks again for this invaluable website.
Nasim, an interesting article. As a commenter above stated, it would be nice to see some Canon comparisons within this group. Any chance you can the add Canon 5DMIII and the Canon 1Dx to this analysis and see how they compare with these Nikons? Thank you. :-)
Tom, please see some of my responses above. We will definitely do our best to include more of Canon gear in the future, but for now it is the question of budget limitations when getting gear for such articles…
Hi Nasim
Images from the D600 and D800E were down-sampled to 16 MP .This 0ne resizing makes grain 2 times bigger on D800.If you will apply like noise redusion first ant resizing after real picture is much better. (i been doing this one test between D1x and D800E).You don’t have to do resising- you have just make pictutes same size to make comparison.
Thanks
Vilmas, you mean makes grain 2 times smaller? Because that’s what down-sampling does – reduce grain size. I did not want to add the noise reduction process, because it is another variable that could affect the data. If you use noise reduction, then it should be applied to all images and it will deviate from true performance results…
Hi Nasim,
Great. I am not a professional photographer; With my D800E/D7000 I have hardly take pictures above ISO 500; In some extreme cases I may go till ISO 1800. Whether professionals really use 6400 & above?
Every time a new camera is out, they keep pushing the ISO to extreme levels. Is it a value add for professionals?
Thanks,
Of course it isn’t ;-), it’s just a marketing gimmick, as Nasim correctely pointed out..
I don’t know about using ISO 6400+, but on my D800 I would regularly and happily use anything over ISO 1000 if conditions warranted. My main subject is aerial photography and I would rather use a high shutter speed to remove any camera movement and let the ISO take care of itself. Blurred images are worth nothing.
Further, I photograph the occasional event where I much prefer to record the ambiance of the room as well as the subjects with a little bit of direct flash to lighten up the eyes. I remember the old “soot and whitewash” days when photographing in large rooms using film with a speed of ISO ((ASA in those days!) of 400 if you were lucky.
If I knew how, I would post a couple of examples to show you what I mean.
Cheers
Sylvester, I often have situations where I do have to go higher on ISO, sometimes above ISO 3200 on my D800. I try to keep it low, but if the situation is difficult, I know that I can get good quality results when the image is resized to a smaller version.
Wow, the d800 is really significantly sharper when downsampled compared to the lower res cameras.
Martin, yes, that’s what down-sampling does and that’s why high resolution sensors have their own advantages when compared to lower resolution sensors. They cannot quite keep up with very high ISOs due to massive loss of colors and dynamic range, but below ISO 12800, they are very good.
Yeah, I just haven’t been around to see it this clearly in an example before. Coincidentally, this is also just a few days before I am myself upgrading to a d800, so I couldnt be happier haha =)
nice compare
all of them are great… im so interested in Df
I was comparing the D600 to sony A7 and I ve found that the test photos
are not constant although it is supposed to have the same ISO, check them
out
cdn.photographylife.com/wp-co…O-6400.jpg
cdn.photographylife.com/wp-co…O-6400.jpg
Nasim, very high ISO like ISO 204,800, is not a marketing gimmick. It can be used for surveillance.