DPReview Studio shot comparison Question...

[DNBush]

I've used this tool to compare how different cameras handle high ISO noise but in my most recent comparison, I've noticed something.

The comparison was between the following cameras with the images all shot at ISO6400. I downloaded the RAW files and processed them using DxO and only applied noise reduction. I made no other adjustments.

OM-1

Canon EOS R6 Mark II

Canon EOS R7

Canon EOS R10

Now, beyond just how noise was handled, what I noticed was that colors differed which might be expected but it's how they differed that I found odd.

The OM-1 and R7 looked almost identical.

The R6 Mark II and R10 also looked almost identical.

Now in the case of the R10 and R6 II, what's similar is that both are 24mp sensors, though of course one is APS-C and one FF. They also used different lenses for the shots.

The R7 and OM-1 are not only different sensors both in size and mp but they also used different lenses. Yet, the colors appear nearly identical.

So I guess my question is, is the studio scene tool, only of value for ISO handling comparisons? What am I to make of this color situation? Are there just too many other factors involved with color that would make a comparison not possible?

 

[Erik Baumgartner]

The color is up to you and how you use your RAW editor. The color can be made to match, but may not with the default processing.

 

[DNBush]

The color is up to you and how you use your RAW editor. The color can be made to match, but may not with the default processing.

True but I made no changes to color in DxO. All images we're processed using identical settings (default).

 

[Erik Baumgartner]

The color is up to you and how you use your RAW editor. The color can be made to match, but may not with the default processing.

True but I made no changes to color in DxO. All images we're processed using identical settings (default).

Raw files from different cameras commonly produce somewhat different color using the same settings in every RAW editor I’ve used. Color profiles aren’t always perfectly matched for every camera.

 

[Quarkcharmed]

I've used this tool to compare how different cameras handle high ISO noise but in my most recent comparison, I've noticed something.

The comparison was between the following cameras with the images all shot at ISO6400. I downloaded the RAW files and processed them using DxO and only applied noise reduction. I made no other adjustments.

OM-1

Canon EOS R6 Mark II

Canon EOS R7

Canon EOS R10

Now, beyond just how noise was handled, what I noticed was that colors differed which might be expected but it's how they differed that I found odd.

The OM-1 and R7 looked almost identical.

The R6 Mark II and R10 also looked almost identical.

Now in the case of the R10 and R6 II, what's similar is that both are 24mp sensors, though of course one is APS-C and one FF. They also used different lenses for the shots.

The R7 and OM-1 are not only different sensors both in size and mp but they also used different lenses. Yet, the colors appear nearly identical.

So I guess my question is, is the studio scene tool, only of value for ISO handling comparisons? What am I to make of this color situation? Are there just too many other factors involved with color that would make a comparison not possible?

Apart from processing discussed above, DPR doesn't guarantee the studio scene itself is exactly the same every time. While illumination may be consistent, the elements of the scene fade out over time - at some point it was renewed a few years ago. There was a mention somewhere to not use the scene for colour comparison.

If you process raw files from the tool yourself, at least try and set the white balance using the colour checker in the middle of the scene.

 

[Bill Ferris]

The DPR Studio Scene can be used to compare noise performerance at selected exposures of up to four different cameras, simultaneously. Those comparisons can me made at the pixel level or the sensor level. It also demonstrates the perceptible resolution disparities of sensors with different megapixel counts. There are a variety of questions one can explore in these areas.

You can also download raw files and JPEGs from each camera to work with in your photo processing and editing app of choice. That exercise can shed light on how you may or may not enjoy working with the files produced by a camera being considered for purchase.

It's as useful a resource as a person wishes to make of it.

 

[DNBush]

The DPR Studio Scene can be used to compare noise performerance at selected exposures of up to four different cameras, simultaneously.

Which is what I have used it for.

Those comparisons can me made at the pixel level or the sensor level. It also demonstrates the perceptible resolution disparities of sensors with different megapixel counts. There are a variety of questions one can explore in these areas.

Can you elaborate on what some of those questions might be and how to go about doing that?

You can also download raw files and JPEGs from each camera to work with in your photo processing and editing app of choice. That exercise can shed light on how you may or may not enjoy working with the files produced by a camera being considered for purchase.

Done that. But, it seems from what I'm hearing, you can't use the tool to explore color differences between camera bodies.

 

[Bill Ferris]

The DPR Studio Scene can be used to compare noise performerance at selected exposures of up to four different cameras, simultaneously.

Which is what I have used it for.

Those comparisons can me made at the pixel level or the sensor level. It also demonstrates the perceptible resolution disparities of sensors with different megapixel counts. There are a variety of questions one can explore in these areas.

Can you elaborate on what some of those questions might be and how to go about doing that?

If you want to see the potential noise benefits of working with a larger sensor at the same exposure, use the "Comp" display. If you want to explore the exposures needed for cameras of different formats to make photos with similar noise visibility, use "Comp" and different ISOs.

You can do the above and compare the performerance of current and older gen cameras. That's useful if you're considering the purchase of a used body. I've seen members post comparisons demonstrating the read noise performance of different format cameras.

If you put your mind to it, there are probably other interesting comparisons to be made.

You can also download raw files and JPEGs from each camera to work with in your photo processing and editing app of choice. That exercise can shed light on how you may or may not enjoy working with the files produced by a camera being considered for purchase.

Done that. But, it seems from what I'm hearing, you can't use the tool to explore color differences between camera bodies.

That depends on what you mean by, color differences. You can certainly take a raw file into an image processing app to explore how much work is required to get the colors looking as you want. If your goal is to see how satisfying colors will be in a straight-out-of-camera JPEG, thats not as clear. The images are intentionally given a rather uniform vanilla processing treatment to avoid that swaying one's evaluation of camera performance.

 

[DNBush]

The images are intentionally given a rather uniform vanilla processing treatment to avoid that swaying one's evaluation of camera performance.

Hmm.... But what's showing up is not uniform at all. That was what got me thinking that the differences I am seeing are due to the sensors.

 

[Bill Ferris]

The images are intentionally given a rather uniform vanilla processing treatment to avoid that swaying one's evaluation of camera performance.

Hmm.... But what's showing up is not uniform at all. That was what got me thinking that the differences I am seeing are due to the sensors.

There are several factors to account for when exploring how different camera raw files appear on a computer screen, especially when opened in a photo editing app. That's how I would recommend you approach a more thorough evacuation: download the raws and open them in your app.

It's difficult to isolate just the sensor in that kind of comparison since the camera and lens perform as a system, including processing that's done in camera. Some cameras apply noise reduction to raw files and that can't be baked out.

There's also the question of how your image processing app demosaics different brand raw files. One obvious example are Fujifilm RAFs vs, for example, Nikon NEFs. That's a situation where Fujifilm's sensor architecture plays an observable role in producing observably different displays.

Is the app configured to apply no preset, a camera-based preset, or a custom preset when demosaicing the raw?

Differences in color hue and saturation that you see when comparing different brand raws made with low exposures and minimal light energy are impacted by shot noise, read noise, chroma noise and possibly pattern noise. The question I would ask is, how well can this be addressed in Lightroom Classic; my image editing app of choice?

If color rendering is a priority for you, I suggest searching the DPR image galleries or using the Google machine to search for manufacturer sample images. The manufacturer has a vested interest in presenting their product in the best possible light. Color rendering is near the top of that list. If you love the color you see in the sample gallery photos for a particular model camera, at least you it's possible to get that look; even if you don't know the secret sauce of in-camera settings that produced it.

 

[ishwanu]

I downloaded the RAW files and processed them using DxO and only applied noise reduction. I made no other adjustments.

Now, beyond just how noise was handled, what I noticed was that colors differed which might be expected but it's how they differed that I found odd.

What am I to make of this color situation?

It should help you realize that cameras do not all produce the same colors, and a RAW converter/editor does not produce the same colors from all cameras when the default parameters are applied, and all available RAW converter/editors are unlikely to produce the same colors even from the same file when their default parameters are applied.

In short ... no, the studio shot comparison tool is not meant for careful color analysis, but it does help demonstrate the challenge of color matching among cameras.

 

[UncleVanya]

DXO isn’t enough info. What DXO product? Photolab? Which version?

Why am I asking? Because over the years DXO has approached things differently. In sine versions the body profile is automatically applied. In some versions you need to apply the DXO standard preset to get this. You can look into the color rendering section and see what’s applied. I’m not a fan of worrying about color differences. Generally the jpg profiles can be tweaked and certainly the raw output can be tweaked. But the amount of work needed to get what you want is a valid consideration.