Colorspace in Canons JPEG (and DPP). Different output

[boldcolors]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

The same phenomena can be observed in DPP. When everything is set to sRGB and the file is exported to Photoshop we get a consistent look. BUT when you change the color space in DPP to Wide Gamut you notice a saturation and contrast boost, export the image to Photoshop and convert it to sRGB. This - again - give another result (more saturated) compared to the all-sRGB flow.

What's up with that? I like it though. It makes me want to use Canon JPEGS more often because with some tweaks in camera the JPEGS can look really really good. Something I have only experienced with Fujifilm before.

Is Canons sRGB more limited than "standard" sRGB?

 

[NAwlins Contrarian]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

IMO the point of shooting JPEGs is to get files you can use (as proofs or otherwise) without any post-processing. If you're going to post-process, then shoot raw (or raw + JPEG). Raw files certainly give you more control and potentially better quality.

IMO 'richer' in this context is at best an imprecise term, but depending on the raw converter, you can certainly increase saturation and/or vividness, and/or decrease lightness or some or all colors.

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

The same phenomena can be observed in DPP. When everything is set to sRGB and the file is exported to Photoshop we get a consistent look. BUT when you change the color space in DPP to Wide Gamut you notice a saturation and contrast boost, export the image to Photoshop and convert it to sRGB. This - again - give another result (more saturated) compared to the all-sRGB flow.

What's up with that?

That mainly depends on how the camera maps the capture to a color space for a JPEG versus how the convert-to-profile is done in post. If you convert from Adobe RGB to sRGB using relative colorimetric rendering intent, then you may well get more saturated colors than if you just shot sRGB JPEGs. Of course, you can also get clipped colors (and/or more clipped colors). TNSTAAFL.

Is Canons sRGB more limited than "standard" sRGB?

The camera's ability to capture colors is far wider than sRGB's ability to contain them. How any particular camera maps the colors it captures to sRGB may well affect what you see.

 

[CAcreeks]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

Could you please post some examples?

Or direct us to the relevant thread at Fred Miranda site?

I don't believe it. Usually colorspace conversions lose data, especially 8-bit ones.

 

[dwalme]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

The same phenomena can be observed in DPP. When everything is set to sRGB and the file is exported to Photoshop we get a consistent look. BUT when you change the color space in DPP to Wide Gamut you notice a saturation and contrast boost, export the image to Photoshop and convert it to sRGB. This - again - give another result (more saturated) compared to the all-sRGB flow.

What's up with that? I like it though. It makes me want to use Canon JPEGS more often because with some tweaks in camera the JPEGS can look really really good. Something I have only experienced with Fujifilm before.

Is Canons sRGB more limited than "standard" sRGB?

Doesn't surprise me at all. The tool used to represent that data as sRGB has to determine what to clip, what to compress, and by how much.


Each tool will do this a little differently. Some tools allow you to configure the end result.

 

[boldcolors]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

Could you please post some examples?

Or direct us to the relevant thread at Fred Miranda site?

I don't believe it. Usually colorspace conversions lose data, especially 8-bit ones.

I think you misunderstand. It seems Canons sRGB space in the camera (and DPP) is more limited than it needs to be. Setting the camera (or DPP) to Adobe RGB enhances colors and they are kept when later converting to sRGB in Photoshop.

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

I really don't know what else to say. This thread was mainly aimed towards Canon users so maybe I should have posted it in another forum. But If you want proof and don't use Canon I can recommend downloading a Canon RAW file here on this site, open it in DPP and set it to sRGB first. Export to Photoshop and keep it there. Then go back to the same image in DPP, set it to Adobe RGB and export it again to Photoshop. This version should be richer than the first one. Now convert the Adobe RGB version to sRGB and compare the two.

 

[NAwlins Contrarian]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

Could you please post some examples?

Or direct us to the relevant thread at Fred Miranda site?

I don't believe it. Usually colorspace conversions lose data, especially 8-bit ones.

I think you misunderstand. It seems Canons sRGB space in the camera (and DPP) is more limited than it needs to be. Setting the camera (or DPP) to Adobe RGB enhances colors and they are kept when later converting to sRGB in Photoshop.

I think you misunderstand. It is fundamental truth that the camera can capture lots of ranges of colors well outside of sRGB. The issue is--if the final output is going to be sRGB--how and at what stage the colors in the image outside of sRGB are dealt with.

There are two basic photography-appropriate models of how to do this, and any given camera an/or software can use either or something intermediate between the two.

(1) With a 'relative colorimetric'-type approach, in-gamut colors are recorded as they are and out-of-gamut colors are clipped to the nearest in-gamut color. This results in overall higher color saturation but loses subtle tonal transitions / tonality in high-saturation areas of the image.

(2) With a 'perceptual'-type approach, the saturation of some or all colors is reduced so that even the most saturated colors in the image are recorded as (just) in-gamut, with less-saturated colors recorded with proportionally lower saturation. This results in overall lower color saturation but retains tonality in high-saturation areas of the image.

One good source of explanation of these processes is Andrew Rodney's article you can freely download from his website. Go to this page and click "Rendering Intents and ICC Profiles".

But basically, if the camera records JPEGs using a 'perceptual' or hybrid process, you have the camera record a JPEG in Adobe RGB color space, and in software you perform a 'convert to profile' to sRGB using a relative colorimetric-type process, then you will get a more saturated image--at the expense of tonal separation in the saturated colors. Again TNSTAAFL.

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I really don't know what else to say. This thread was mainly aimed towards Canon users so maybe I should have posted it in another forum. But If you want proof and don't use Canon I can recommend downloading a Canon RAW file here on this site, open it in DPP and set it to sRGB first. Export to Photoshop and keep it there. Then go back to the same image in DPP, set it to Adobe RGB and export it again to Photoshop. This version should be richer than the first one. Now convert the Adobe RGB version to sRGB and compare the two.

Some of us may do that, and for the reasons explained above, the same issue / tradeoff may apply to other cameras. And FWIW, I think this is a perfectly fine place to raise the issue.

 

[CAcreeks]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

 

[boldcolors]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

 

[CAcreeks]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

But what is the point of converting back to sRGB?

Using the most recent versions, every browser in the world is color managed and can display Adobe RGB just the way you like it.

Printing might be a problem, but nobody does that any more.

 

[boldcolors]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

But what is the point of converting back to sRGB?

Using the most recent versions, every browser in the world is color managed and can display Adobe RGB just the way you like it.

Printing might be a problem, but nobody does that any more.

A lot of bigger companies/clients are using older software and have strict upgrade policies so I still believe sRGB is the safest choice even in 2024

 

[Nabilon]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

How do you define ‚richer colors‘? Is it possible that the image converted from Adobe-rgb to s-rgb contains more clipped colors and thus appears more saturated? That would also explain the higher contrast, as the values are spread out over the whole range. I would guess that the amount of information in both images is the same, but spread out differently. The s-rgb image might contain more details in the shadows and highlights, while the Adobe-rgb image loses those details during the conversion to s-rgb, but maintains more details in the midrange instead. It all depends on the algorithm of conversion. The difference is similar to softproofing with either perceptual or relative intend.

 

[boldcolors]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

How do you define ‚richer colors‘? Is it possible that the image converted from Adobe-rgb to s-rgb contains more clipped colors and thus appears more saturated? That would also explain the higher contrast, as the values are spread out over the whole range. I would guess that the amount of information in both images is the same, but spread out differently. The s-rgb image might contain more details in the shadows and highlights, while the Adobe-rgb image loses those details during the conversion to s-rgb, but maintains more details in the midrange instead. It all depends on the algorithm of conversion. The difference is similar to softproofing with either perceptual or relative intend.

Richer as in more saturated and I highly doubt clipped. Here is an example by another person over at FredMiranda

Canon-mount SLRs - Photography - FM Forums

Explore Canon-mount SLRs photography discussions. Share your photos, discuss gear and techniques, and buy or sell equipment in our marketplace.

www.fredmiranda.com

It's a standard headshot in perfectly balanced daylight/studiolight. As you can see the sRGB version is less saturated (in this case though I would actually say more accurate) compared to AdobeRGB. The Adobe RGB has also a different curve which could explain the saturation increase BUT after adding the same curve to the sRGB image, the AdobeRGB one is still more saturated.

Here is another example with dominant reds which of course is always a clipping-risk but it doesn't look like that.

Canon-mount SLRs - Photography - FM Forums

Explore Canon-mount SLRs photography discussions. Share your photos, discuss gear and techniques, and buy or sell equipment in our marketplace.

www.fredmiranda.com

And to repeat what I have written earlier - if I convert an AdobeRGB Canon file in Photoshop to sRGB the color saturation and contrast is usually still the same (regardless choice of intent) so that would confirm that colors are not clipped.

My only takeaway (again) is that Canons own approach to sRGB is a more "compressed" conversion - if that makes sense.

For what it's worth, when exporting the image from DPP set to sRGB the embedded profile is called sRGB v1.31 (Canon) (16bpc). So it's not the regular sRGB profile for whatever reason.

 

[Zeee]

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I found a thread on Fred Miranda's site, maybe this is it. However I see very little difference. Maybe it's an issue of somebody using a non color managed browser?

https://www.fredmiranda.com/forum/topic/1790888/0

Not sure which is which, but the second one is darker and a bit redder. DPP bug?

If I set Firefox gfx.color_management.mode = 0, the difference is much more visible.

Yep. And to my eyes the second one is significantly warmer and has higher saturation as well as a different curve/contrast.

But what is the point of converting back to sRGB?

Using the most recent versions, every browser in the world is color managed and can display Adobe RGB just the way you like it.

Printing might be a problem, but nobody does that any more.

Intersting. I have not printed in years and only do so if needed. In my elder years just shooting, processing, posting at photo forums is enough for. I can always export using LrC and choose media type type when I shot events. I export several sets of files. Two sets for glossy and matte printing. Several sizes of screen viewing, one just for previewing. All numbered identically so customers could quickly pick what they wanted to print or post.

I'm a little off topic but in the last 6 months I have been using Adobe Neutral which I prefer over Canon Camera Neutral. Huge difference when to comes to working with whites/highlights and blacks/shadows.

https://www.alex-kunz.com/linear-profiles-refutation/

 

[Ab S]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

Could you please post some examples?

Or direct us to the relevant thread at Fred Miranda site?

I don't believe it. Usually colorspace conversions lose data, especially 8-bit ones.

I think you misunderstand. It seems Canons sRGB space in the camera (and DPP) is more limited than it needs to be. Setting the camera (or DPP) to Adobe RGB enhances colors and they are kept when later converting to sRGB in Photoshop.

I think you misunderstand. It is fundamental truth that the camera can capture lots of ranges of colors well outside of sRGB. The issue is--if the final output is going to be sRGB--how and at what stage the colors in the image outside of sRGB are dealt with.

There are two basic photography-appropriate models of how to do this, and any given camera an/or software can use either or something intermediate between the two.

(1) With a 'relative colorimetric'-type approach, in-gamut colors are recorded as they are and out-of-gamut colors are clipped to the nearest in-gamut color. This results in overall higher color saturation but loses subtle tonal transitions / tonality in high-saturation areas of the image.

(2) With a 'perceptual'-type approach, the saturation of some or all colors is reduced so that even the most saturated colors in the image are recorded as (just) in-gamut, with less-saturated colors recorded with proportionally lower saturation. This results in overall lower color saturation but retains tonality in high-saturation areas of the image.

One good source of explanation of these processes is Andrew Rodney's article you can freely download from his website. Go to this page and click "Rendering Intents and ICC Profiles".

But basically, if the camera records JPEGs using a 'perceptual' or hybrid process, you have the camera record a JPEG in Adobe RGB color space, and in software you perform a 'convert to profile' to sRGB using a relative colorimetric-type process, then you will get a more saturated image--at the expense of tonal separation in the saturated colors. Again TNSTAAFL.

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I really don't know what else to say. This thread was mainly aimed towards Canon users so maybe I should have posted it in another forum. But If you want proof and don't use Canon I can recommend downloading a Canon RAW file here on this site, open it in DPP and set it to sRGB first. Export to Photoshop and keep it there. Then go back to the same image in DPP, set it to Adobe RGB and export it again to Photoshop. This version should be richer than the first one. Now convert the Adobe RGB version to sRGB and compare the two.

Some of us may do that, and for the reasons explained above, the same issue / tradeoff may apply to other cameras. And FWIW, I think this is a perfectly fine place to raise the issue.

Just for my understanding because I only shoot RAW. I always assumed that when shooting RAW the camera setting of color space (sRGB or AdobeRGB) has no influence. Only the RAW processor determines the color space (in my case DxO PL7 Wide Gamut) applied in the converted Jpg?

 

[boldcolors]

Due to an ongoing discussion over at Fred Miranda about color science in Canons JPEGs there was a claim that one can get richer sRGB colors in Canons JPEGS by setting the color space in camera to Adobe RGB and then convert back to sRGB in post rather than setting it to sRGB default in camera (if one wants to stay in sRGB space that is)

I did some tests and it really works. Which is puzzling. The Adobe RGB version converted back to sRGB in Photoshop has deeper, slightly more vivid colors than the same image taken with sRGB set in the camera from the beginning. Of course, same image/jpeg settings in both scenarios. Only changing color space in camera.

Could you please post some examples?

Or direct us to the relevant thread at Fred Miranda site?

I don't believe it. Usually colorspace conversions lose data, especially 8-bit ones.

I think you misunderstand. It seems Canons sRGB space in the camera (and DPP) is more limited than it needs to be. Setting the camera (or DPP) to Adobe RGB enhances colors and they are kept when later converting to sRGB in Photoshop.

I think you misunderstand. It is fundamental truth that the camera can capture lots of ranges of colors well outside of sRGB. The issue is--if the final output is going to be sRGB--how and at what stage the colors in the image outside of sRGB are dealt with.

There are two basic photography-appropriate models of how to do this, and any given camera an/or software can use either or something intermediate between the two.

(1) With a 'relative colorimetric'-type approach, in-gamut colors are recorded as they are and out-of-gamut colors are clipped to the nearest in-gamut color. This results in overall higher color saturation but loses subtle tonal transitions / tonality in high-saturation areas of the image.

(2) With a 'perceptual'-type approach, the saturation of some or all colors is reduced so that even the most saturated colors in the image are recorded as (just) in-gamut, with less-saturated colors recorded with proportionally lower saturation. This results in overall lower color saturation but retains tonality in high-saturation areas of the image.

One good source of explanation of these processes is Andrew Rodney's article you can freely download from his website. Go to this page and click "Rendering Intents and ICC Profiles".

But basically, if the camera records JPEGs using a 'perceptual' or hybrid process, you have the camera record a JPEG in Adobe RGB color space, and in software you perform a 'convert to profile' to sRGB using a relative colorimetric-type process, then you will get a more saturated image--at the expense of tonal separation in the saturated colors. Again TNSTAAFL.

As for examples, in this specific case If I were you and didn't believe a poster then I would want to test it myself and not ask for already processed images in a workflow I didn't know about. Or are you saying that I am not seeing what I am seeing?

IMO those making claims based on their eyeball observations should post sample files so others can make their own eyeball evaluations and discuss. And of course, some people will then subject the files to objective analysis and post the results, which is also helpful.

I really don't know what else to say. This thread was mainly aimed towards Canon users so maybe I should have posted it in another forum. But If you want proof and don't use Canon I can recommend downloading a Canon RAW file here on this site, open it in DPP and set it to sRGB first. Export to Photoshop and keep it there. Then go back to the same image in DPP, set it to Adobe RGB and export it again to Photoshop. This version should be richer than the first one. Now convert the Adobe RGB version to sRGB and compare the two.

Some of us may do that, and for the reasons explained above, the same issue / tradeoff may apply to other cameras. And FWIW, I think this is a perfectly fine place to raise the issue.

Just for my understanding because I only shoot RAW. I always assumed that when shooting RAW the camera setting of color space (sRGB or AdobeRGB) has no influence. Only the RAW processor determines the color space (in my case DxO PL7 Wide Gamut) applied in the converted Jpg?

Correct but that is not what we are discussing

 

[NAwlins Contrarian]

Just for my understanding because I only shoot RAW. I always assumed that when shooting RAW the camera setting of color space (sRGB or AdobeRGB) has no influence. Only the RAW processor determines the color space (in my case DxO PL7 Wide Gamut) applied in the converted Jpg?

When you shoot raw (or raw + JPEG, as it pertains to raw), the color space setting on the camera, or sRGB or Adobe RGB or whatever, is mostly just a tag. In the raw converter (like you I use DxO PL7 with wide gamut) you can process however and export in whatever color space you want.

However, the color space setting may have a subtle (probably very subtle) effect insofar as it may slightly affect metering and/or indication of clipping (highlights or shadows). These may be calculated relative to the 'JPEG' version of the image. Even raw-only sticks a JPEG thumbnail in the raw file. So e.g. an area of the image that would have colors clipped a bit in sRGB may not have those colors clipped in Adobe RGB, which might prompt the camera to choose (in auto-exposure) and/or show you (clipping warnings) a slightly different exposure.

FWIW I shoot raw + JPEG and set my cameras to sRGB because I want the JPEGs to be 'least common denominator' proofs / immediate-distribution versions.

 

[CAcreeks]

Intersting. I have not printed in years and only do so if needed. In my elder years just shooting, processing, posting at photo forums is enough for. I can always export using LrC and choose media type type when I shot events. I export several sets of files. Two sets for glossy and matte printing. Several sizes of screen viewing, one just for previewing. All numbered identically so customers could quickly pick what they wanted to print or post.

Interesting! Did you (or LrC) create different renditions for glossy vs matte prints?

When we had more wall-space to fill with photos, I printed on glossy inkjet coated paper, and results were good. Epson inks never lasted, but Canon inks lasted for years. I tried matte inkjet paper but didn't love it.

I'm a little off topic but in the last 6 months I have been using Adobe Neutral which I prefer over Canon Camera Neutral. Huge difference when to comes to working with whites/highlights and blacks/shadows.

https://www.alex-kunz.com/linear-profiles-refutation/

Useful essay, but I prefer S curves.

I'll bow out of the main thread because saturation can be increased in virtually any software, or on a monitor, so switching colorspaces seems like a waste of time for me. Also in my view, Canon DPP colorspace conversion is buggy. With a color managed browser, Adobe RGB and sRGB should look nearly identical (more than the do in the Fred Miranda posts).

 

[Ab S]

Thanks for the confirmation and advice (apologies for interrupting the original discussion)!

 

[Zeee]

Intersting. I have not printed in years and only do so if needed. In my elder years just shooting, processing, posting at photo forums is enough for. I can always export using LrC and choose media type type when I shot events. I export several sets of files. Two sets for glossy and matte printing. Several sizes of screen viewing, one just for previewing. All numbered identically so customers could quickly pick what they wanted to print or post.

Interesting! Did you (or LrC) create different renditions for glossy vs matte prints?

Yes. Ink absorbs into matte stock and not on gloss. The size of the printed dot will be different to compensate.

When we had more wall-space to fill with photos, I printed on glossy inkjet coated paper, and results were good. Epson inks never lasted, but Canon inks lasted for years. I tried matte inkjet paper but didn't love it.

I'm a little off topic but in the last 6 months I have been using Adobe Neutral which I prefer over Canon Camera Neutral. Huge difference when to comes to working with whites/highlights and blacks/shadows.

https://www.alex-kunz.com/linear-profiles-refutation/

Useful essay, but I prefer S curves.

I'll bow out of the main thread because saturation can be increased in virtually any software, or on a monitor, so switching colorspaces seems like a waste of time for me. Also in my view, Canon DPP colorspace conversion is buggy. With a color managed browser, Adobe RGB and sRGB should look nearly identical (more than the do in the Fred Miranda posts).