Adding Custom Working Profiles

Keentolearn · May 13, 2020, 6:22pm

Hi Roel

No bullshit coming from you for sure…and i said way back that I want to learn.

What I thought Holmes was genuine about was that his choice of 5 working spaces had:

A. The ability for more chromacity to be added to each of his variable sized spaces (with variants )without making the space itself bigger!. i have no idea how that is done or how it works . i took his advice and used DCam 3

B Holmes sells the idea that “large “ Pro Photo spaces have to shrink in size to fit into a smaller printer space with possible uncomfortable results.Again I would not have your knowledge to contradict him

Is there better than prophoto from a printing perspective?

Read Elles reports…awesome…but with the following comment…

“The ProPhotoRGB primaries are hard-coded into Adobe products such as Lightroom and the Dng-DCP camera “profiles”. However, other than being large enough to hold a lot of colors, ProPhotoRGB has no particular merit as an RGB working space. Personally I recommend the Rec.2020 or ACEScg profiles over ProPhotoRGB. But if you have an already well-established workflow using ProPhotoRGB, you might find a shift to another RGB working space a little odd, at least at first, and so you have to weighthe pros and cons of changing your workflow.”

Do i go for Rec 2020 or Aces cg or ProPhotoRGB if i move away from Holmes and will i get a smooth transition to Imageprint…

Thanks for a pile of help…

Matt

Claes · May 13, 2020, 6:38pm

@Keentolearn

Is there better than prophoto from a printing perspective?

Prophoto is far too good when it comes to printing.
Here is a small gamut comparison I made some time ago:

Have fun!
Claes in Lund, Sweden

Keentolearn · May 13, 2020, 6:37pm

“Since you are using RT for upsampling now, perhaps consider using it for your raw processing too.”
I certainly will

“use embedded”
Makes sense in case one forgets to do the manual switch

“Do TIFFs enjoy 32 bit floating point precision for editing in RT if they are 15 bits to start with (coming out of Photoshop)?”

This is the sort of information i really would appreciate having …thanks for raising the question Samuel

Matt

Keentolearn · May 13, 2020, 6:45pm

Thank you Afre
thank you Desmis
Matt

afre · May 13, 2020, 6:58pm

@jdc When I pinged you, I neglected to specify the reason. Wanted you to comment on the merits of changing the working profile and the merits or demerits of doing so. Thanks for pitching in!

Keentolearn · May 13, 2020, 7:06pm

Your right on it…
Its gotta be me
I will practice gap free coding tonight and conquer the monster…
Be back in the morning with goods news…good night and thanks
Ps you are incredibly patient!
Matt

Thanatomanic · May 13, 2020, 7:07pm

@Keentolearn Your working profile should be bigger than your output profile, and your output profile should be chosen for your intended medium. If you publish for the web, this is sRGB. If you print, this is device dependent. You can usually obtain ICC profiles from the better copyshops that match their printers. If you transfer your image to another software for post-processing, it should be as large as possible (easiest would be to match your working profile).

Like @Claes mentions, the gamut of ProPhoto is much wider than that of any ordinary printer. Therefore, at the end of the processing pipeline, the gamut is mapped (shrunk) to the output profile. Usually the printer gamut is also smaller than that of your screen, so if you choose your output profile well, you can preview the image as it would turn out on your paper. (There’s also a thing called soft-proofing, you can look that up if you like to know more).

A. The ability for more chromacity to be added to each of his variable sized spaces (with variants )without making the space itself bigger!.

I think that in essence increasing chromaticity is nothing more than pushing colors towards the boundaries of the gamut. It would seem logical that the bigger your gamut is, the further you can push. So having different ‘flavors’ doesn’t make sense to me. To see the effect in RawTherapee, try playing with the L*a*b* tool and push the chromaticity slider to a large value and then change your working profile. That should give a noticeable change in color.

“Do TIFFs enjoy 32 bit floating point precision for editing in RT if they are 15 bits to start with (coming out of Photoshop)?”

Yes. All internal maths is done in floating points.

And finally regarding uncomfortable results, and different results when switching between working profiles. Imagine you are decorating a room and have put all the furniture exactly where you want. Then you suddenly change the size and angles of your walls. Your furniture needs adapting. That is how a change in working profile works as well.

XavAL · May 13, 2020, 7:08pm

I will just add a plot of all 3 color spaces discussed here:

Blue: ProPhoto
Green: Rec2020
Red: DCam3

Note: DCam 5 is just ACES p0

Edit: somehow the plots in the previous image got shifted. Fixed now

Keentolearn · May 13, 2020, 10:34pm

Wanted you to comment on the merits of changing the working profile and the merits or demerits of doing
For years ACR did not offer the choices of working spaces like you can get for instance in RT now.
Then later Holmes was included in ACR if you wished it to be included right from after the camera NEF enters ACR and across to and through Photoshop.
I know very little about the technical side of profiling other than to watch gamut warning (with the Imageprint print profile as the space to gamut watch) when i am working on a file in Photoshop…i found the fancy retina screen was misleadingly optimistic in terms of a preview of the printed finished product …I know now that RT has no problem with me starting in ACR/ Holmes then across to Photoshop/Holmes …then on to RT/Holmes and finally back to Printing in Photoshop/Holmes / Imageprint if i want to combine the best of both programs.I’m the problem in that I am a complete dunce when it comes to getting the json file set up.
I trusted Joseph Holmes opinion that in terms of landscape photography that DCam3 was able to handle natures colours very well running on to and through a decent printer profile and that a bigger working colour space would require a compression clash (ok Perceptual does its bit to soften that ) to come down to gamut size for print. DCam 3 obviates ,in his opinion, this compression at print time . I don’t want clipping if it is caused by moving in and out of bigger/smaller spaces.You guys have the knowledge to do your own smooth processing without having to rely on a packaged product …i didn’t…i will experiment with some of the spaces in Rt now for sure.
thanks
matt

Keentolearn · May 13, 2020, 10:54pm

(There’s also a thing called soft-proofing, you can look that up if you like to know more).

You didnt notice that I mentioned earlier on that I have been using a RIP (Imageprint) for years now.
I have profiles from Imageprint for my papers and of course the printer itself.I never work other than with softproofing turned on.
I also use their Phatte Black system which ,as you know , allows one to print a quality b/w print or a good Colour print without flushing inks.
Yes. All internal maths is done in floating points.
Great to know…I’m learning a lot today…

Keentolearn · May 13, 2020, 11:35pm

The Colour spaces map is exactly what i expected to see…super
Adobe RGB and Srgb would be smaller again than DCam3. Would you have an RT printer profile to correspond roughly in size to DCam3 ? Something in the RtV4 range would be a winner or would RT V2 be more realistic ?
I would like to lie low in terms of green( for printing) if at all possible

samuelchia · May 14, 2020, 1:53am

I would say merits would include:

using the smallest possible space to contain the image’s colours to minimise wastage of gamut
working on non-raw images that are in a colour space not included in the bundled working profiles list, and not wanting an internal conversion to happen (though I don’t think there will be any material damage of rounding of numbers since RT uses 32 bit floating point for its calculations, as long as the working profile chosen can encompass the input profile’s gamut entirely)

Demerits would include

behaviour of some sliders would change depending on the working profile chosen, which may make the “feel” of them less intuitive if swapping between many different working profiles constantly
choosing a too-small working profile may result in clipping issues if not careful

And perhaps this might help, from the RawPedia Color Management Addon page:

What happens when we are in a “Working Profile” or when we change of “Working Profile” or the settings?

When we are in a narrow profile like sRGB, it may seem obvious that the colors will be limited to this profile limits! So, if an initial color (the sensor one, slightly modified by interpolation) is inside sRGB, what happens when we tweak sliders and curves?:

in RGB mode (Exposure), if saturation or lightness or contrast is/are modified, we work in rgb mode or its linear derived “hsv”. This means that the resulting effects will depend on the working space ((sRGB, AdobeRGB, Prophoto), the results will be different when we move from “sRGB” to “Prophoto”, even if we stay inside the gamut limits.

in Lab mode, if lightness, chromaticity, contrast or curves are modified, we modify directly the L, a, b, (or C,h) values. This means that if we stay inside the gamut limits, the image will be the same (within the limits of the color management)

When exceeding the gamut - that is, the original image - either acting the sliders or the curves, what happens?

let’s take an example, an original color is L=27 a=2 b=-75; this color is in the Prophoto space and is worth R=42 G=52 B=158, and in sRGB R=-85 G=69 B=184 (negative R means out of gamut). If we change the working profile, it is obvious this color can’t be restored; the XYZ conversion will give L=32 a=21 b=-67 and R=0 G=69 B=184. The color will have a different aspect because it is completely impossible to reproduce it in a smaller color space.

second example, a color inside sRGB : L=40 a=42 b=-44, that is in RGB (sRGB) R=133 G=66 B=166, and in RGB(Prophoto) R=102 G=64 B=140: a) if we apply saturation (+30) in exposure, the values become - Prophoto - L=36 a=49 b=-49 and sRGB L=38 a=47 b=-47, because we act on the RGB values; b) if we apply chromaticity in Lab adjustments; Lab values become L=40 a=55 b=-58 as well in sRGB than Prophoto, because we stay inside the sRGB gamut; c) if we choose a color, close enough to the sRGB gamut limits, but inside the gamut: L=40 a=63 b=37 and we apply chromaticity +30, Lab values become - Prophoto L=40 a=81 b=49 - we can notice that the hue is preserved (arctg(b,a)), and in sRGB L=44 a=69 b=50 – the hue isn’t preserved.

samuelchia · May 14, 2020, 2:02am

@Thanatomanic

Just to clear some things up. BTW I’m not advocating for Joe’s profiles nor am I trying to sell them to anyone.

In Elle Stone’s writings, ProPhoto was not highly praised in relation to other profiles like ACES and Rec.2020. I’m sorry that the writing style has put you off but there’s some basis to Joe’s claims.
Adobe does not sell their profiles. It is not unusual to put testimonials for things you sell. No comment on the pricing of the profiles, he is free to price as he wishes and you are free to decide if you want to purchase or not.
The Dcam spaces (these are the working spaces) do not intend to determine any particular look and feel of any image. The comments of chroma modification only refer to the chroma variants of the Dcam profiles and this effect is essentially the same as the chromaticity slider in RawTherapee in the Lab module. The “better chroma” is “better” in the sense that if you want to increase the colourfulness of your image, you are better off with such an adjustment than the typical saturation slider which honestly is pretty awful.

samuelchia · May 14, 2020, 2:17am

Actually the chroma variants for increasing colourfulness are bigger than Dcam3 to be able to work. They are progressively uniformly larger versions of Dcam3 (different primaries) which are to be assigned and NOT converted to for the effect to take place. It is like changing the units and not the numbers. 1 cm → 1m. By this change of units, “1” is now physically a greater distance. Similarly, the chroma variants that decrease colourfulness are progressively uniformly smaller versions of Dcam3.

BTW the LUTs of DCP camera profiles are limited to Prophoto primaries, if I understand Anders Torger correctly. I’m not entirely sure what meaningful differences there will be in using ACEScg over Prophoto.

afre · May 14, 2020, 6:22am

@Keentolearn @samuelchia I am not looking for an answer per se but rather if there is any specific our friend wants to say about replacing the default working profile with another.

Correct me if I am wrong: I am getting the impression that we are becoming preoccupied with the terms and concepts rather than the practical implications of them. My advice is start playing with raw files and make actual comparisons. PlayRaw is the perfect way to do that.

gaaned92 · May 14, 2020, 10:43am

Be aware that some linear transformation are not independant of the working space in which they are performed, mainly when there are multiplications.

For instance, if you use channel mixer (matrix multiplication) in different WS, you will get different results.

Interesting. So the values out of prophoto gamut could be clipped if using a DCP?

Does’t it mean that there should be a way to assign (and not use) a color profile to the generated output?
It seems also mean that DCAM3 is an output profile?

samuelchia · May 14, 2020, 1:29pm

Will be clipped. Exactly. The DCP must be an LUT profile specifically for this limitation to occur. Which is also why I’m uncertain about how advantageous it would be to switch to ACEScg. I doubt it will make any material difference even to a fussy regular photographer.

It was intended to be used in Photoshop where assigning profiles can be easily done.

Dcam3 is a working space RGB profile like sRGB, Adobe RGB, Prophoto etc. and I don’t mean “Working Profile” in the case of RawTherapee. Indeed it would be what you can use as the “Output Profile” in RT, although I normally think of output profiles as device profiles like display and printer.

jdc · May 14, 2020, 2:37pm

@afre
Vast debate where everyone has their point of view. It’s like now everyone is a virus specialist and an epidemiologist

I think, Prophoto, is not the best, because blue Y is near of zero. With this choice some sliders are almost without action.
In addition these colors (deep blues with very low luminance) are barely visible or invisible
Widegamut, ACES_AP1, Rec2020 seems good enough

Thanatomanic · May 14, 2020, 4:27pm

There is still a lot of confusion here.

The input profile is the only profile that you assign in the literal sense: you choose how you want to interpret the input pixel values. Is your source encoded in sRGB? Then assigning an sRGB profile as your input profile is the only reasonable thing to do. Assign a different profile and your image won’t look right. Is your source a camera raw file? Then you need to assign a camera specific ICC or DCP profile to get your colors right.

The input profile converts input values to the XYZ reference color space, and then, usually, immediately to your working profile. There is no way to ‘assign’ this profile. The working profile is a large color space in which you can safely do your (non)linear image editing. A large space, like ProPhoto or ACES AP0, gives you plenty of room for (extreme) edits to your image, without causing color clipping or banding (assuming you work with floating points). Importantly, many image operations (edit: not all) require a linear profile (i.e. TRC with gamma = 1.0). Using a different gamma in your working profile will cause unexpected or undesired effects. All working color spaces in RawTherapee force a gamma of 1.0 (in other words, you only choose the primaries).

There are some caveats like @gaaned92 and @jdc point out: some tools are linked to the shape and size of the color space. That is simply how some operations need to be performed. The choice of primaries therefore influences how these tools work. That is why changing your working profile after you finished an edit, is generally a bad idea and may change your colors significantly. Simply pick a large enough color space when you start and then don’t worry about it anymore. Chances are you will never encounter any issues while editing.

Finally, you can choose an output profile. This profile should be determined by your intended medium. If you go for print, pick a printer profile. If you go for the web, pick sRGB. When there are differences in the gamut of your output and working color spaces, some gamut mapping needs to take place. There are multiple ways to do this (perceptual, colorimetric, …), which all slightly change the final look of your image. But this is the only function of the output profile.

Then onto the DCam profiles. If the profile is large enough, then sure, you can use that profile in place of ProPhoto as a working profile.
Holmes’ claims about “better perceptual linearity” with his “proprietary tone curve” for a working profile is complete nonsense. That is not the point of a working profile at all. You may want to do that for an output profile, but then the question is: why do you need such a large gamut in case of DCam3/4/5? Your intended medium probably doesn’t cover it.

So my main claim here is: anything that applies a specific look to your image (deliberate changes in chromaticity, adding of a tone curve) should never be done through any of these profiles. Unfortunately Holmes disagrees with me

XavAL · May 14, 2020, 5:18pm

Or any other wide gamut, for the sake of completeness.

There’s one instance where it is the way to go: when you use a raw processor as an intermediate application to perform non-raw edits (as wished by @Keentolearn), or as a starting point for your image, that will be further processed in other programs before arriving to the final image.