Dynamic range is the ratio between the strongest and weakest signal measured in a given context. For example, the dynamic range of a camera sensor is the ratio between the brightest and darkest signal that can be captured in a single image. Like most other things in photography, dynamic range is often measured in stops, with one stop brighter meaning twice as much light.
Although dynamic range can be used in any context where there is a signal to be measured – such as sound – I’ll be talking about dynamic range in photography.
Table of Contents
Dynamic Range of the Scene
The dynamic range of a scene is the ratio between the brightest and the dimmest parts of a scene. Of course, using the term “the scene” is a bit vague. One way to define it more precisely would be to say the dynamic range of a scene is the dynamic range of the scene projected onto a surface through an optical system. Think of the projected image onto a sensor, rather than what is recorded by the sensor itself.
The dynamic range of a scene may be any value, and it could even be larger than the dynamic range of the device with which you are trying to record the scene.
Dynamic Range of Our Eyes
The dynamic range of the human eye is a bit more complex. Unlike a camera, the eye does not simply take a snapshot of the image projected on the retina and store it in the brain for viewing. Instead, the brain also performs complex and powerful processing that ends up as what we consciously perceive. In effect, the brain continuously reads the signals from the eye, and it integrates this information into a single cohesive whole.
By looking at different areas of a scene, you can perceive a fairly large dynamic range. Over time, the human eye can actually adapt to different brightness levels in the range of about 46.5 stops. However, part of that range consists of dark scenes can that only be sensed by the rods which are not sensitive to color.
The simultaneous dynamic range or steady-state dynamic range on the other hand is basically the dynamic range that can be perceived at any one time without any adaptation. The measurement of this quantity is rather interesting because there has been some disagreement about it.
For example, earlier research claimed 6.6 stops, and this is essentially the figure quoted on Wikipedia. However, this figure was derived from an experiment that measured the time taken to adapt to rapid color changes. In other words, this original figure was not obtained from an experiment designed specifically to measure dynamic range. This figure was also reported in a book by Robert L. Myers which may explain why it is so often cited.
Researchers at the University of Bristol actually designed a proper experiment to directly measure the simultaneous dynamic range of the eye, and they obtained an average value of 12.4 stops.
Dynamic Range of Cameras
At last, we come to the most important thing in life: the dynamic range of your camera sensor. This is the ratio between the maximum signal that can be captured by the sensor and the noise floor of the sensor. The noise floor is essentially the point at which a part of the image would be indistinguishable from an image taken with no light hitting the sensor at all. Photographic enlightenment!
Sometimes people use different measures of dynamic range since dark parts of the image can still be higher than this noise floor while still being fairly useless photographically. However, these minutiae are beyond the scope of this article.
DXOMark is one website that publishes dynamic range measurements of cameras. For example, here are some measurements at the base ISO of each camera:
Full-frame camera | Dynamic range (stops) | APS-C camera | Dynamic Range (stops) | Micro 4/3rds camera | Dynamic Range (stops) |
---|---|---|---|---|---|
Canon R5 | 14.6 | Canon 77D | 13.3 | Panasonic GH5II | 13.1 |
Sony A7IV | 14.7 | Sony A6600 | 13.4 | Olympus OM-D E-M1 Mark II | 12.8 |
Nikon Z7II | 14.7 | Nikon D500 | 14.0 | Olympus Pen-F | 12.4 |
Pentax K1 | 14.6 | Pentax K3 II | 13.6 | Panasonic GH4 | 12.8 |
The dynamic range of your camera depends on the ISO setting. For example, the Sony A7IV as measured by DXOMark has a dynamic range of 14.7 stops at base ISO but only 10.2 stops at ISO 6400. Typically, the highest dynamic range is obtained at the base (non-extended) ISO.
We also see that dynamic range increases with sensor size. Therefore, it will be easier to capture the dynamic range of a dramatic landscape with a larger-sensor camera with just one shot, which is one of many reasons why full-frame cameras are often preferred for landscapes.
In my opinion, these numbers are definitely not as useful as they used to be. For example, at one time there was a large difference between Canon and Nikon in terms of dynamic range. Take the Nikon D610 and Canon 5D Mark III, which were both released at about the same time. The D610 has a dynamic range of 14.4 at base ISO according to DXOMark, whereas the 5D Mark III has a dynamic range of 11.7. This difference of 2.7 stops could actually be noticeable in real-world comparisons, although of course great images can be made with both cameras.
These days, any current full-frame cameras will give you some healthy number over 14 stops. Moreover, there are so many other factors affecting image quality that sensor dynamic range should not be overemphasized.
What You Should Know About Dynamic Range
Knowing that your camera has 14.3 stops versus 14.8 stops probably won’t help you… unless maybe you use your camera’s dynamic range as a lottery number and you win. However, there are a few things about dynamic range that are helpful. The first is that when you are presented with a very high dynamic range scene, you need to try and fit as much of that into your final image as possible.
For digital cameras and a single exposure, that means you should expose to the right; in other words, you should expose so that the brightest important parts of your image are as bright as possible without being overexposed. Moreover, for landscapes, you should do this at base ISO (except when you can’t).
With film, the situation is a little different. Overexposure is more easily recovered with negative film, whereas shadows detail is very easily lost. This is the opposite of digital, where highlights are easily lost but shadow detail is easily recovered.
Therefore with negative film, it makes sense to overexpose a little to get the additional shadow detail. This does not apply to positive or slide film which loses detail in overexposure like digital.
Finally, with mostly static scenes such as many landscapes, taking multiple exposures and combining them using something like the averaged high dynamic range technique will also increase dynamic range. For a scene with a lot of dynamic range, dark areas that would otherwise be too noisy can be captured in such an averaged image, without overexposing the highlights.
Conclusion
Dynamic range is nothing mysterious. It simply tells you how great the difference is between the brightest and darkest signals, either in a scene or perceivable by a sensor. In some types of photography such as landscape photography, the dynamic range of scenes is often high, and therefore it’s important to fit as much of it in your final image as possible. Luckily, modern cameras are so good that you should be able to get great images even in challenging conditions.
Great explanation. Thanks
I want to get your weekly email to get more great content like this!
I have a D610, Z6, Z6ii and as of 2 days ago, a Z9 which I have yet to edit any photos with. My photography partner has a Z6, D850 and Z9. He was always raving about the dynamic range of the D850. I didn’t get understand because the dynamic range stats didn’t suggest much difference. But when I edited some of the photos from his D850, I couldn’t believe how much detail could be recovered from highlights that appeared way overblown.
I’m somewhat confused by what the theory tells me about the dynamic range and my real world experience recovering details from highlights with the D850.
For nature photography I use my D810 with the 180-400 mm lens. Do I still need a D500? My XT3 serves me well as a tourist camera when I want to travel light. Who in your opinion has the fastest tracking, sharpest focus system in a camera/lens combination that’s not too heavy?
Why when looking at a person with the sun to their back do I see a better light impression of not just that person but the whole scene as compared to taking a photo where properly exposing for that subject overexposes the rest of the photo? Is one’s mind so fast at combining everything that it all registers as one scene?
Yes
“steady-state dynamic range”
Love it! As opposed to “dynamic dynamic range” of course 😀
I actually did not realize that was oxymoronic until you pointed it out!
NB: The following is not criticism aimed at Jason Polak; it is intended to be useful information.
After performing some web searches, it seems to me that far too many authors (Jason excluded) have a deep‑seated need to replace well‑known terms with more science‑y sounding terms, which is, of course, the realm of pseudoscience.
I think the old-fashioned technical term for “steady‑state dynamic range” was contrast.
In the fields of both analog and digital signal processing, I used the technical term dynamic range to mean the same as that described in the Wikipedia article:
“Dynamic range … is the ratio between the largest and smallest values that a certain quantity can assume.”
That wording is very important: “can assume” NOT “does assume” NOR “assumes”. Specifically, it is the maximum range, or the capability, of the complete system.
I suggest that a scene cannot have a dynamic range because:
• it is not part of the imaging system per se;
• parts of the scene can be so dark that no photons reach the image sensor.
• the scene is an actual ontic entity, which has actual photons, but it does not contain abstract epistemic quantities such as intensity ranges.
ontic: what is
epistemic: what we know
I agree! Thanks for the comments. Two points: the terms I used for the dynamic range without adaptation are just the ones from the eye/medical research papers, which seems logical so that people can find credible information from primary sources. And second, your final point is why I defined the “dynamic range of the scene” to be the dynamic range of the “signal” of the projection onto a surface, so that there is a well-defined quantity to be measured.