My color editing recommendations for 2024 - (new: DxO Wide Gamut everywhere)

[Wandering Photographer]

Advertisement (gone after registration)

23 hours ago, mzbe said:

Even if you throw away much of the gamut due to output constraints, at that point you still win if you have preserved information along the way.

I used to print flowers in B&W whenever I could not preserve the details and texture I cared about on a color medium - perhaps you have similar experiences?

I have photographed flowers and seen the texture I see in the flower disappear in the sRBG image. I typically save images in both raw format and sRGB. If I look at it in raw, I frequently see the texture become visible as long as I don’t get clipping in any of the primary colors.

If the scene fills a color gamut wider than sRBG, then if your raw file records those as distinct values outside the sRBG gamut, compressing mapping the pixels to the sRBG gamut will irretrievably destroy information because you will be mapping multiple values to one. If you save it to JPEG, you lose more information because of lossy compression. Depending on what your goal is as a photographer this may or may not be important.

 

23 hours ago, mzbe said:

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Wandering Photographer]

20 hours ago, hansvons said:

The working colour space in today's photo editing applications colour space, which usually is linear (there are exceptions like ACEScct which has a logarithmic response curve and is often used in video colour grading), is larger than the camera’s and human sight’s working space and much, much larger than display-referred colour spaces like sRGB or P3.  Thus, there is no reason for trying to stay close to a small colour space with a gamma response curve such as sRGB. 

On the contrary, sRGB, P3, even Adobe98 (which is not display-referred and often used for printing due to its compatibility) are never a good idea for a working space as their comparable small gamut visually cripples the much larger gamut of camera raw data (after debayering to RGB).

The only reason why one would choose sRGB, P3, or Adobe98 as a working space (if the application would allow that) would be the editing of JPEG files that were exported in the respective colour space. However, the transform LUTs that transform sRGB/Adobe98 to linear (that's what the editing application uses) are fairly robust and simple, without any noticeable translation losses as its math is invertedly based on the usual output transforms we use for exporting our pictures to sRGB making them compatible to 99% of the worlds displays (that’s why sRGB is display-referred and not scene-referred as are the colour spaces of our raw cameras). 

This raises a question. For both my Nikon and Leica digital cameras, I am offered the choice of either sRGB or Adobe RGB. I always specify the Adobe RBG. I save both the raw and the JPEG. When I look at the raw image in Adobe Bridge it says it is in Adobe RGB. How do I access a wider gamut than Adobe RGB? Is that just to provide a wider pallet for manipulation in Photoshop? If so how do I actually see this in the image if my display shows 98% of the Adobe RGB? Then of course there is the gamut of limitations of my printer (Epson SureColor P700) and paper.

[Guest]

2 hours ago, Wandering Photographer said:

This raises a question. For both my Nikon and Leica digital cameras, I am offered the choice of either sRGB or Adobe RGB. I always specify the Adobe RBG. I save both the raw and the JPEG. When I look at the raw image in Adobe Bridge it says it is in Adobe RGB. How do I access a wider gamut than Adobe RGB? Is that just to provide a wider pallet for manipulation in Photoshop? If so how do I actually see this in the image if my display shows 98% of the Adobe RGB? Then of course there is the gamut of limitations of my printer (Epson SureColor P700) and paper.

The sRGB and Adobe RGB settings affect the JPEG image only. The Raw data in the DNG always represents the full native gamut of the sensor - some of it outside sRGB and Adobe RGB.

[jaapv]

2 hours ago, Wandering Photographer said:

This raises a question. For both my Nikon and Leica digital cameras, I am offered the choice of either sRGB or Adobe RGB. I always specify the Adobe RBG. I save both the raw and the JPEG. When I look at the raw image in Adobe Bridge it says it is in Adobe RGB. How do I access a wider gamut than Adobe RGB? Is that just to provide a wider pallet for manipulation in Photoshop? If so how do I actually see this in the image if my display shows 98% of the Adobe RGB? Then of course there is the gamut of limitations of my printer (Epson SureColor P700) and paper.

When converting yourDNG in ACR you can change the settings in the blue link at the bottom of the screen. Set to 16 bits and at least Adobe RGB.  My recoomendation is to use Adobe RGB if you have an sRGB monitor and Prophoto if you have an Adobe RGB or P3 one. 

[hansvons]

6 hours ago, Wandering Photographer said:

This raises a question. For both my Nikon and Leica digital cameras, I am offered the choice of either sRGB or Adobe RGB. I always specify the Adobe RBG.

If you use Adobe RGB/98 in the camera, you will have JPEGs that are not 100% compatible with the rest of the world (there will be colour shifts) as the web, people's monitors, phones and whatnot are calibrated to sRGB. That's why sRGB makes more sense for most applications unless you want to edit/colour-correct the Adobe RGB/98 files and export them as sRGB final images. However, grading off JPEGs regardless of what colour space they represent makes no sense when a proper RAW file is available.

[pgk]

I always work on the following principles:

1. You cannot reproduce the colours you see.

2. In 99.999% of photography colour will be subjective and absolute colour fidelity is not only impossible, it is also usually irrelevant. Viewers will judge whether they like the colour balance, not wheter it is precisely accurate.

3. Most people view images on their monitors with the monitors too bright.

So I'm a bit empirical in that if the image I see on the monitor prints pretty well then I'm happy. sRGB is not the best colour space but it is the most used so we are prettywell stuck with it most of the time.

Lastly, if I'm undertaking a colour critical shoot then I will use a colour checker chart. This is usually when copying .....

As a last thought, if you need the widest colour gamut space to edit in then it might be worth considering why. Are you overediting?

Edited January 4, 2024 by pgk

[jaapv]

Advertisement (gone after registration)

9 hours ago, pgk said:

As a last thought, if you need the widest colour gamut space to edit in then it might be worth considering why. Are you overediting?

No it is because I want to see as much  as possible of the colours that I have available, whether I am editing or not. My monitor is Adobe RGB, my printer accepts Adobe RGB, why should I use a smaller space? Better one step larger to give some elbow room especially when converting back from LAB. My subjects tend to be landscape and wildlife where greens are important. As it happens green has the most differentiation  gain in the larger colour spaces and blue (skies) as well Costs nothing, brings benefits. 
Of course I will dumb down to sRGB for web use. 

[hansvons]

9 hours ago, jaapv said:

No it is because I want to see as much  as possible of the colours that I have available, whether I am editing or not. My monitor is Adobe RGB, my printer accepts Adobe RGB, why should I use a smaller space? Better one step larger to give some elbow room especially when converting back from LAB. My subjects tend to be landscape and wildlife where greens are important. As it happens green has the most differentiation  gain in the larger colour spaces and blue (skies) as well Costs nothing, brings benefits. 
Of course I will dumb down to sRGB for web use. 

Spot on. 

9 hours ago, jaapv said:

why should I use a smaller space? Better one step larger to give some elbow room especially when converting back from LAB.

Would you mind elaborating briefly on where LAB conversions come into play in your workflow? Thanks!!

[jaapv]

Well, that requires more bandwidth than this forum allows. I use LAB for colour manipulation, especially if locally without masking - Use blending layers and split the sliders. Also for enhancing  colour contrast and detail by compressing the A and B curves. Controlling contrast without influencing colour saturation in the L curve comes to mind as well.

The best  I can do is recommend the books by Dan Margulis  " Modern Photoshop Color Workflow" and especially "Photoshop Lab Color, the Canyon Conundrum" Warning: They come with a headache and I do not pretend to understand them completely.

[pgk]

9 hours ago, jaapv said:

No it is because I want to see as much  as possible of the colours that I have available, whether I am editing or not. My monitor is Adobe RGB, my printer accepts Adobe RGB, why should I use a smaller space? Better one step larger to give some elbow room especially when converting back from LAB. My subjects tend to be landscape and wildlife where greens are important. As it happens green has the most differentiation  gain in the larger colour spaces and blue (skies) as well Costs nothing, brings benefits. 
Of course I will dumb down to sRGB for web use. 

The thing is, that I have only seen colour problems when an image is initially poor (exposure/colour balance/colour bias) in 'normal' photography. Personally  I edit in (PS) Adobe RGB then convert to sRGB for output (even for publishers these days as its what they are supplied with, and failing to convert an Adobe RGB file will result in a flat looking final image (it happens)). I am very dubious whether operating in any larger colour space has any distinguishable advantage for general photography. Best practice is good practice but there is reality too.

[hansvons]

4 minutes ago, jaapv said:

Well, that requires more bandwidth than this forum allows. I use LAB for colour manipulation, especially if locally without masking - Use blending layers and split the sliders. Also for enhancing  colour contrast and detail by compressing the A and B curves. Controlling contrast without influencing colour saturation in the L curve comes to mind as well.

Thanks. Thought so. So you are deeply buried in that rabbit hole 🙃. Perhaps you might want to take a look at 3D LUT Creator. This program offers you a wealth of options and will send you even further down (probably you already have it). I stopped exploring LAB, RGBW, and other colour concepts and returned to classic colour timing based on RGB (especially on film scans). Too time-consuming. I use the same workflow for still photography regarding white AND black balancing giving me balanced neutrals from black to white and a solid foundation for further colour editing.

Neither LR nor C1 has a dedicated tool for proper black balancing like you have in proper video grading software. In C1, you can work around that with the Levels tool set to RGB in single-channel mode. I think PS offers a similar option. 

 

[hansvons]

21 minutes ago, pgk said:

I am very dubious whether operating in any larger colour space has any distinguishable advantage for general photography.

Yes, there will be. Your camera's colour space is larger than Adobe RGB 1998. When importing the camera data into Adobe RGB, information will be lost.

That's why, in cinematography, the Academy of Motion Pictures introduced ACES a decade ago, a digital linear colour space offering over 32 stops of latitude and a gamut that is larger than the human eye can see and, thus, eternally future-proved. Today, it's the de facto standard for film grading, offering enough "space" for any cine camera.

LR's ProPhoto working colour space is of a similar size and significantly larger than Adobe RGB, so is C1's own proprietary 32-bit linear colour space. 

Edited January 5, 2024 by hansvons

[pgk]

4 minutes ago, hansvons said:

Yes, there will be. Your camera's colour space is larger than Adobe RGB 1998. When importing the camera data into Adobe RGB, information will be lost.

That's why, in cinematography, the Academy of Motion Pictures introduced ACES a decade ago, a digital linear colour space offering over 32 stops of latitude and a gamut that is larger than the human eye can see and, thus, eternally future-proved. Today, it's the de facto standard for film grading, offering enough "space" for any cine camera.

LR's ProPhoto working colour space is of a similar size and significantly larger than Adobe RGB, so is C1's own proprietary 32-bit linear colour space. 

Yes that's all great, but cinematography does require a consistency rarely required in 'general' photography. I suppose that my point is that its all very well adopting precision and best practice and suchlike, but .... . And the but is that for 'general' photography it is very possible to get excellent results adopting simpler practices. All too often we concentrate on technicalities which were (and are) in the realms of precision technical photography without actually checking to see whether they are either necessary or relevant, or indeed, whether the end results are actually distinguishably better. I pity many taking up photography today because the complexities suggested and adopted are often of little if any practical advantage. I am all for 'best practice' but I am also very aware that reality is quite a harsh critic and avantages need to be observable.

I am technically critical but my workflow is such that it is very straightforward and I don't obsess over things unless they need to be colour accurate. As it happens I do do some work which require colour consistency above all else and even here the workflow is best not complicated because any complication risks future problems if something changes.

As an aside I wonder if HCB would have shot JPEG or RAW files?

[jaapv]

It is not only colour consistency. It is colour differentiation which will result in visibly better detail. Even after converting down to sRGB. 

[hansvons]

2 minutes ago, pgk said:

All too often we concentrate on technicalities which were (and are) in the realms of precision technical photography without actually checking to see whether they are either necessary or relevant, or indeed, whether the end results are actually distinguishably better.

Can't agree more. From a creative, artistic point of view, 95% of what this forum discusses is irrelevant to the audience of one's work. It just happens to be a technical forum. For relief, I can't recommend this thread enough: 

 

[hansvons]

1 minute ago, jaapv said:

It is colour differentiation which will result in visibly better detail. 

Absolutely. In terms of colour, it's all about colour differentiation. Consistent neutrals from black to grey to white precisely provide that. 

[jaapv]

HCB was l finicky about the darkroom work of his printer. He supplied exceedingly detailed instructions.

[jaapv]

Just now, hansvons said:

Absolutely. In terms of colour, it's all about colour differentiation. Consistent neutrals from black to grey to white precisely provide that. 

And that is where LAB comes in by intensifying colour contrast without impacting luminance. 

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[jaapv]

8 minutes ago, jaapv said:

And that is where LAB comes in by intensifying colour contrast without impacting luminance. 

Example: if you shoot a lawn it will be a more or less amorphous green. If you go into LAB and narrow the A and  B curves without shifting the center point the overall colour will remain the same but you will see the individual blades of grass (resolution permitting) when you shift back to RGB and desaturate slightly. 

[Wandering Photographer]

On 1/5/2024 at 7:06 AM, jaapv said:

Example: if you shoot a lawn it will be a more or less amorphous green. If you go into LAB and narrow the A and  B curves without shifting the center point the overall colour will remain the same but you will see the individual blades of grass (resolution permitting) when you shift back to RGB and desaturate slightly. 

I have switched my default color space in Photoshop from Adobe RGB to ProPhoto RGB. I will let you know if I notice any difference in the end result in a few months. This has been a useful discussion.