fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Sony Forum | Join Upload & Sell
1      
2
       end
   

Utility of dynamic range beyond a certain point
  
 
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #1 · Utility of dynamic range beyond a certain point


old-gregg wrote:
Everyone is excited about the increased dynamic range of the Sony A7RVI at nominal ISO. Meanwhile, I am asking myself whether I ever bump into the DR limitations of my existing cameras in practice?

When I look at my favorite landscape shots, for example, I see that - while there's plenty of unused room to lift shadows - lifting them even further results in unrealistic "HDR" look. I think you all are familiar with that look. In other words, at least in my work, there's a natural limit of how much the luminance range of a scene can be "compressed" into...Show more →

I think that continuing to improve dynamic range is a fine thing, and I’m glad to see companies continuing to do that. The improvements since early digital cameras have been really useful. Back then I often had to use techniques like exposure bracketing to deal with wide-DR scenes.

On the other hand, I see a lot of people making a much bigger deal of incremental improvements than seems appropriate.

I like to frame the issue this way, in part: There are basically three situations you’ll run into relative to dynamic range…

1. The great majority of photographs don’t challenge the dynamic range of any camera and DR isn’t really an issue at all.

2. The dynamic range of the scene is so huge (imagine shooting straight into the sun AND wanting clear details on the shadowed side of a very dark rock) that no camera has sufficient dynamic range to capture it all. You either need to do exposure blending and/or accept some blown highlights (sometimes OK) or blocked shadows (also sometimes OK). This is, in my experience, the second most common situation, though it is far less likely than #1 above.

3. The dynamic range of the shot barely exceeds the capabilities of one camera but lies wholly within the range of an alternative camera. Imagine that your 12-stop system isn’t wide enough but your 13-stop system is. This is the least likely scenario but far… and since the DR limit is not a “hard stop” but instead a rolling off of acceptable quality, there are usually things that one can do to deal with it.

Regarding #3, recent significant improvements in things like noise reduction have also increased the range that we can coax out of our cameras. I have a 10-year-old high MP camera that is, by the standards of many, very outdated when it comes to DR. Yet I only rarely ever need to resort to exposure bracketing any more since I can recover/maintain so much image quality in post.

A case in point. While a A7rVI appears to have improved DR capabilities over the A7rV, almost no one is going to get enough benefit from that in real world photography to warrant tossing the older camera and replacing it with the new one. (There might be other reasons for a few photographers…)
Jun 10, 2026 at 03:18 AM
johnvanr
Offline
• • • • •
Upload & Sell: On
p.2 #2 · Utility of dynamic range beyond a certain point


Surfnsun wrote:
The only time I run into dynamic range issues (speaking generally) is when I set up the shot poorly. It doesn’t matter whether I’m dealing with shadows or highlights. Most of the time, it’s not the camera’s fault. Between the cameras I already have, I honestly don’t think I have much use for anything with more dynamic range than what I’m getting right now.

Exactly. More dynamic range can fix my mistakes. But usually, I dislike too much of it.


Jun 10, 2026 at 04:52 AM
guidostow
Offline
• • • •
Upload & Sell: Off
p.2 #3 · Utility of dynamic range beyond a certain point


For woodland photography I absolutely would like more DR. But since I don't need any of the autofocus/faster sensor read/video geegaws sony added to the a7rivi, I'm wondering if a Fujifilm GFX 100s might be a better way to get there... For much of my current work I'm using converted canon TS-E lenses anyway, they convert to the GFX just as easily...
Jun 10, 2026 at 06:25 AM
ruthenium
Online
• • • •
Upload & Sell: On
p.2 #4 · Utility of dynamic range beyond a certain point


snapsy wrote:
Shadow vs highlight in DR is a metering/exposure construct, or at least can be normalized as such. For example, if sensor "A" has 2x the saturation capacity vs sensor "B" (ie, 1EV greater DR near saturation) and sensor "B" has 1/2 the read noise vs sensor "A" (ie, 1EV greater DR near shadow clipping), then both sensors can be exposed to capture the same selection and number of a given scene's highlights, in this case by exposing sensor "B" wth 1EV less exposure vs sensor "A".

I am interested in this subject, perhaps because my understanding is shallow.
In your example, the dynamic range of sensor "A" at the base ISO is calculated from the ratio 2S/N.
The dynamic range of sensor "B" is calculated from the ratio S/(0.5N), where the S and N take the same values in both calculations.
The resulting DR is the same for these sensors.
Assuming the same sensor surface area, the exposure on "B" must be one-half (-1 EV) of that on sensor "A".

A case of interest to me is that of my GFX100S II (PDR 12.48 @ISO 80) compared to the new Sony A7RVI (PDR 12.55 @ISO 100) and A7RV (PDR 11.69 @ISO 100).

Since the sensors in the GFX100S II and A7RV are both made by Sony and have the same pixel pitch, it is reasonable to assume that the noise (N) is the same in the calculation leading to the PDR values for these cameras.
Then the difference in the PDR (12.48 - 11.69 = 0.79) is due to a larger S for the GFX. The sensor of this camera can capture close to 1 stop more light at ISO80, compared to the amount of light that the A7RV can capture at ISO100.

Now, we can continue with the PDR 12.55 of the A7RVI. And here, I face a situation that is confusing. The PDR is larger than that for A7RV (11.69), yet this is not achieved by capturing more light with the A7RVI. To the best of my understanding, this better PDR of A7RVI is the result of applying selectively a lower gain toward the "shadow signal" from the sensor. In fact, this is similar to the above sensor "B" where the signal-to-noise ratio was S/(0.5N). Note that this is a pure digital manipulation. I believe that the intrinsic camera noise is similar in both A7RV and A7RVI.

My question is whether achieving a better PDR by capturing more light, or by applying a selective low gain toward the shadow signal should give equivalent images (equivalent, as in photographic equivalence).
My feeling is that there must be some differences between the two methods of increasing the dynamic range.
I would appreciate any comments on this.
Jun 10, 2026 at 06:44 AM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.2 #5 · Utility of dynamic range beyond a certain point


ruthenium wrote:
I am interested in this subject, perhaps because my understanding is shallow.
In your example, the dynamic range of sensor "A" at the base ISO is calculated from the ratio 2S/N.
The dynamic range of sensor "B" is calculated from the ratio S/(0.5N), where the S and N take the same values in both calculations.
The resulting DR is the same for these sensors.
Assuming the same sensor surface area, the exposure on "B" must be one-half (-1 EV) of that on sensor "A".

A case of interest to me is that of my GFX100S II (PDR 12.48 @ISO 80) compared to the...Show more →

Absolutely correct. Here is a comparison of the input-referred read noise for each of the above cameras you mentioned:

https://tinyurl.com/5n9ym4jp

ruthenium wrote:
My question is whether achieving a better PDR by capturing more light, or by applying a selective low gain toward the shadow signal should give equivalent images (equivalent, as in photographic equivalence).
My feeling is that there must be some differences between the two methods of increasing the dynamic range.
I would appreciate any comments on this.

While the images will have equivalent engineering DR and near-equivalent PDR (the former having a SNR cutoff in the shadows of 1:1, the latter 20:1), the image from the sensor with the larger capacity will have over higher overall SNR (ie, less image noise), owing to the more light captured, all things being close to equal, including quantum efficiency and fill-factor.

Since DR is measured at hard cut-offs (SNR at shadows, saturation at highlights), you can have two sensors with equal DR but significantly-different SNR / overall noise performance.
Jun 10, 2026 at 07:11 AM
ruthenium
Online
• • • •
Upload & Sell: On
p.2 #6 · Utility of dynamic range beyond a certain point


@snapsy - thank you! I appreciate your reply.
(I am familiar with the differences in how the engineering DR and PDR are calculated).
Jun 10, 2026 at 07:18 AM
149113
Offline
• • •
Upload & Sell: On
p.2 #7 · Utility of dynamic range beyond a certain point


Examples are probably needed. This is a shot I had earlier in the year. It was taken at ISO 12,800 and 1/3200. That was really the only combination that could achieve what I was looking to capture: small bird, feeding behavior, fast action and low light. A very tough use case. Much more difficult than swallows to photograph in flight. It's not a shot that has a ton of dynamic range given the ISO but it works well for this scene. In some situations I think DR is a little overstated as must have feature for wildlife. Landscape, I can see the need for greater DR but you also have a little more flexibility for SS and aperture too



If there is a better Prairie Warbler shot taken, I'd like to see it


Jun 10, 2026 at 08:21 AM
jojib
Offline
• • •
Upload & Sell: Off
p.2 #8 · Utility of dynamic range beyond a certain point


I shoot mostly portraits and only do landscapes when I'm on vacation. I'm very happy with the dynamic range of my A7V. On post I can extract more details from the highlights more than any of my cameras. In this image I can raise the shadows more but I like it this way. If my A7V was a basketball player---it'll be Steph Curry---who can basically shoot near half court with ease.
Jun 10, 2026 at 08:24 AM
 


Search in Used Dept. 

chiron
Offline
• • • • •
Upload & Sell: On
p.2 #9 · Utility of dynamic range beyond a certain point


ruthenium wrote:
It is possible that paintings don't have a "dynamic range". First, the DR is calculated from the signal to noise ratio. Paintings don't have noise in the same sense as the electronic sensors. Even if there is something akin to "noise" in a painting (that I don’t know) I am not sure how this can be meaningfully quantified.
Second, the light is not represented naturally in painting. The naturally bright sun or the moon, or midday sky are painted at reduced brightness. Human eyes and eyes of animals and birds, apparently, do not register the absolute brightness - only the...Show more →

Another way of describing dynamic range that is more useful for comparing photos and paintings is:

"the range of brightness levels — from the darkest usable shadow to the brightest usable highlight — that a camera can capture in a single exposure while retaining detail. It’s often measured in stops (or exposure values), where each stop represents a doubling of light. For example, 12 stops means the brightest tone is 2¹² (about 4,096) times brighter than the darkest tone with visible detail" (Richard Butler, dpreview.com).

or, put differently:

"the ratio between the brightest and darkest signal that can be captured in a single image" (Jason Polak, Photography Life).

Thus, paintings and photos can both be understood to have a dynamic range (usually called "tonal range" or "value range" in painting), albeit they are created by different means--a sensor vs. paint and canvas (or other substrate). My very inexpert sense of this at the moment is that oil paintings may have more steps of dynamic range than photos in darker areas (with paintings thus able to produce more dark and dark middle tones) but photos are able to produce more and much brighter highlights and more bright middle tones. Correspondingly, paintings have a much more limited contrast ratio than do photographs, which are capable of showing much more contrast, especially in HDR images.

It also may be that many paintings use a more limited set of their inherent dynamic range, with markedly less overall contrast, while photos usually use more of their dynamic range and thus usually show higher contrast.

The visual effects of these variations in tonalities and contrast produce different visual effects as they are seen by a viewer.

The only approximation to HDR images that I see from paintings are monstrosities like the fluorescent Elvis paintings and others like them that are virtually joke images.

Yet, obviously, many paintings are spectacularly beautiful without doing anything like HDR and with a more limited dynamic and contrast range. That might teach us something about how much dynamic range we actually need to make a beautiful photograph.
Jun 10, 2026 at 09:29 AM
sbay
Offline
• •
Upload & Sell: Off
p.2 #10 · Utility of dynamic range beyond a certain point


chiron wrote:
Yet, obviously, many paintings are spectacularly beautiful without doing anything like HDR and with a more limited dynamic and contrast range. That might teach us something about how much dynamic range we actually need to make a beautiful photograph.

When you say dynamic range, are you talking about the dynamic range of the painting itself? or the scene that the painter captured?

A plein air painter will paint an outdoor scene and have to deal with the exact same dynamic range challenges that a photographer would have to handle. Except the painter relies on their eyes to "capture" the information in the scene and then translate it to the more limited dynamic range of the canvas. A photographer would have to rely on the sensor to do the same, and then during editing map it to the more limited dynamic range of the monitor, or paper.

For paper, people normally talk about contrast range. A matt paper might have a contrast ratio (darkest black to brightest white) of 100:1 and a glossy paper might have a ratio of 200:1. I guess that's roughly to 7 to 8 stops.

I suspect paintings have greater contrast ratio than prints because they can use a wider variety of pigments. Apparently there are specialty paints, like Vanta Black, that can absorb 99.965% of light. I think that works out to a 3000:1 contrast ratio (11.5 stops). But even using something like Vanta Black, they are still mapping the scene they see with their eyes (maybe 15stops?) into the much reduced range of the canvas and paint. So I guess you could say they are still doing "HDR" except manually.
Jun 10, 2026 at 09:48 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #11 · Utility of dynamic range beyond a certain point


Painters — at least old-school painters — do something similar to what our visual system does, though it a slightly different way. (And photographic printers do this same thing.)

When we look at a scene with a huge dynamic range our eyes adjust to accommodate the focus of our attention. When we send our attention to something darker the pupils open ad the whole scene lightens, though we don’t really typically register that change. The opposite happens when our gaze goes to a very bright area — our pupils contract and the entire scene darkens.

Painters and photographers actually don’t want to nor do they need to produce prints/paintings with the full real world dynamic range. In fact, we typically lighten the dark areas of the painting/print and darken the lightest areas so that when one takes in the full image there is little need to visually adjust.

You can do an experiment to see how this works. Some evening, perhaps around early twilight, look at a scene with a nearby division between something dark — the near side of a building, silhouetted trees, etc — and something very bright such as a section of the sky. Look at the dark areas for a moment and you’ll see details in those shadows. Now move your gaze to the sky and pay careful attention to what happens — a fraction of a second later the sky will visibly darken.

sbay wrote:
When you say dynamic range, are you talking about the dynamic range of the painting itself? or the scene that the painter captured?

A plein air painter will paint an outdoor scene and have to deal with the exact same dynamic range challenges that a photographer would have to handle. Except the painter relies on their eyes to "capture" the information in the scene and then translate it to the more limited dynamic range of the canvas. A photographer would have to rely on the sensor to do the same, and then during editing map it to the more limited dynamic...Show more →
Jun 10, 2026 at 10:33 AM
guidostow
Offline
• • • •
Upload & Sell: Off
p.2 #12 · Utility of dynamic range beyond a certain point


Most adults can see 24-25 stops. Matt paper can deliver 4-7.5 stops, glossy 6-8, jpeg 8-11 HDR ~17ish (last two are device dependent). Recording as many stops of noise free data as possible provides the most control over mapping shadow and highlight detail into the chosen output range.
Jun 10, 2026 at 10:59 AM
chiron
Offline
• • • • •
Upload & Sell: On
p.2 #13 · Utility of dynamic range beyond a certain point


sbay wrote:
When you say dynamic range, are you talking about the dynamic range of the painting itself? or the scene that the painter captured?

A plein air painter will paint an outdoor scene and have to deal with the exact same dynamic range challenges that a photographer would have to handle. Except the painter relies on their eyes to "capture" the information in the scene and then translate it to the more limited dynamic range of the canvas. A photographer would have to rely on the sensor to do the same, and then during editing map it to the more limited dynamic...Show more →

I am definitely talking about the dynamic or tonal range of the painting itself. Here's what Claude 4.8 came up with, which is somewhat different from your figures and closer to my experience of viewing images (and I think closer to Dan's points). But I appreciate what you are saying and I know that I may be not understanding your points clearly.

From Claude 4.8:

I asked: "Do oil paintings have anything equivalent to dynamic range in photography?

--Claude answered:--

"Yes, though the equivalent is more constrained and works differently.

"The closest concept is tonal range or the value scale an oil painting can render. A photograph captured digitally might encode 10–14 stops of dynamic range, but any reflective print—oil paintings included—is bounded by the physics of pigment and surface. The brightest possible value is the reflectance of the lightest paint (titanium white, lead white) and the darkest is whatever the darkest pigment absorbs short of perfect blackness. In practice this gives a contrast ratio of roughly 30:1 to 50:1 for most paint surfaces—comparable to a matte photographic print, and dramatically less than what a backlit screen or the original scene offers. So oil paint has a "dynamic range," but it's a narrow one, and the painter's entire craft of value management is essentially the problem of compressing a high-range world into that limited band."

--End of Claude--


What is interesting about painting's more limited dynamic-tonal and contrast range is how painters use it to create beautiful light and images. For example, Vermeer's rendering of light is exceptionally beautiful, yet the dynamic range in his images is very limited.




Edited on Jun 10, 2026 at 05:41 PM · View previous versions
Jun 10, 2026 at 11:00 AM
racetratr
Offline
• •
Upload & Sell: Off
p.2 #14 · Utility of dynamic range beyond a certain point


I rarely need more dynamic range from my camera when shooting at base ISO. Yeah, I like having more leeway in post production, but that's not absolutely critical. When I really could use more dynamic range is when doing fast-paced street photography at night at ISO 6400. Unfortunately, that isn't gonna happen, unless and until there are major reductions in sensor/camera noise.
Jun 10, 2026 at 11:51 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #15 · Utility of dynamic range beyond a certain point


This gets at a common misconception about increasing the dynamic range of cameras. The misconception is that the object is to create a final image with the same dynamic range as the original subject, preferably straight out of camera.

But that’s not the purpose or goal of increasing DR in cameras. It is to capture scene data that spans a wider range of luminosity values without losing highlight details or getting excessive noise in shadow areas…

… so that the image can be PROCESSED to manage the range of dark and light values without introducing too much noise.

And that processing? In general it is a version of what photographers and photographic printers have long done with techniques like dodging and burning —increase the brightness of shadows and control the highlights so as to maintain details there. That is accomplished by DECREASING the dynamic range of the final presentation — e.g. a print.

There’s an analog in audio recording, where we use systems that record dynamic and frequency ranges well beyond what we can actually hear… so that we have the freedom to manipulate those data in audio post without the signals degrading in ways that become audible. But your audio recording that you listen to — whether a digital format or an analog format — no longer contains all of that original data nor can it.

Again, increasing the capability of cameras to record large dynamic ranges accurately is a good thing and something we can continue to look forward to. But it is important to understand what this does and does not actually accomplish and how it is best used.

chiron wrote:
I am definitely talking about the dynamic or tonal range of the painting itself. Here's what Claude 4.8 came up with, which is different from your figures and much closer to my experience of viewing images (and I think closer to Dan's points).

From Claude 4.8:

I asked: "Do oil paintings have anything equivalent to dynamic range in photography?

Claude answered:

Yes, though the equivalent is more constrained and works differently.

"The closest concept is tonal range or the value scale an oil painting can render. A photograph captured digitally might encode 10–14 stops of dynamic range, but any reflective print—oil paintings included—is bounded...Show more →
Jun 10, 2026 at 11:55 AM
DWOfPaul
Offline
• • •
Upload & Sell: Off
p.2 #16 · Utility of dynamic range beyond a certain point


I don't think comparing it to painting is a fair comparison. Often in photography, you have no control over the dynamic range in the photo you are taking. I haven't seen many paintings where the sky is blown out white.

Personally, I am happy to see dynamic range finally expanding. The majority of my landscape / cityscape images I use bracketing on. One stop isn't going to allow me to give up bracketing yet, but it's a step in the right direction. It will also allow more flexibility for wildlife photos where bracketing is rarely an option.

Also, some lenses have a very high level of vignetting. So an extra stop of dynamic range effectively gives you a one stop advantage of noise in the corners for properly exposed images. For example, say your camera has 5 good stops of dynamic range. If your lens has 3 stops of vignetting, you really only have 2 stops of dynamic range to play with after vignetting correction.


Jun 10, 2026 at 04:40 PM
1      
2
       end
FM Forums | Sony Forum | Join Upload & Sell

1      
2
       end
    
 

Welcome back
Log in to your account
Reset passwordNo account? Register