How to determine ICC profile gamut volume

[Bob Rockefeller]

I've seen discussed here in a few threads the concept of the gamut volume of a profile as a way to understand the likely gamut volume of the paper it is used for.

I have a couple of papers I'd like to compare on something other than a purely subjective basis. How can I determine the gamut volume of several profiles I have?

[digitaldog]

I have a couple of papers I'd like to compare on something other than a purely subjective basis. How can I determine the gamut volume of several profiles I have?

ColorThink Pro has a 'gamut volume' metric it reports. That said, I'm not certain if what it reports is what you desire 
I asked the fine folks at CHROMIX what this value provides, here's what I heard back:

The ColorThink Grapher calculates the gamut volume in terms of cubic Lab values.  A Lab value of one is one delta E (dE76 considering the way the Grapher is currently made).  So each of these cubic Lab values represent the smallest discernible color difference, and each cubic Lab value represents a unique human-discernable color.  So in that sense, the larger gamut will necessarily represent more distinct, humanly-perceptable colors than a smaller gamut.   (Now that statement comes with our usual caveat that this volume number is a rough estimate, not a precise one - and it works well for and is intended for making comparisons between profiles, not for defining absolute volume numbers.)

ON the OTHER hand…..

There is a philosophical issue at stake here:  Just what constitutes a color?  Depending on how the numbers are encoded, you can have several million combinations of different numbers representing different colors in theory, but some will point out that these different number combinations do not constitute individual “colors” since they are not distinct enough to be *different* to the human vision system.

- gamut volumes in ColorThink are calculated using color values and refer to the number of unique color values
- unique color values in ColorThink refer to the number of distinct colors (as per human vision) that exist in a color gamut or a color image or color list or whatever.
- In color gamuts we just use cubic Lab values, the idea being that one delta-E76 value away from a color *should* result in a perceptually different color

[Mark D Segal]

Once you've compared the gamut volumes in CTP you will of course have comparative statistics, but what those statistics mean in terms of the comparative reproduction of real-word photographs is another matter - depends very much on the gamut of the image being printed.

[Bob Rockefeller]

All that sounds as if it provides at least a common reference from which to make some comparisons.

I have ColorThink 2.3, which doesn't seem able to calculate a gamut volume. ColorThink Pro is a pretty expensive upgrade.

Are there other tools to make the calculation? Or perhaps reference web sites that have the volumes for a number of papers?

[digitaldog]

PatchTool can:

http://www.babelcolor.com/#PatchTool

Here lies the rub. ColorThink Pro reports Adobe RGB as having a gamut volume of 1,207,520.

[AlterEgo]

or



or apple one

[Iliah]

Using average dE error of a profile as the basis for gamut computation instead of 1 dE and discarding everything with dE larger than 6 gives a better perspective, and fo a good printer with a good profile may result in volume numbers 2 orders of magnitude lower.

[Bob Rockefeller]

Here's an interesting site I saw reference to somewhere or other:

http://www.iccview.de

It's in German, but is easy enough to figure out most of.

It provides comparative 3D gamut diagrams, with calculated gamut volumes, from uploaded ICC profiles.. That volume is reported in "cubic color units." MOAB's profile for their Juniper Baryta Rag is calculated at 8,125,647 cubic color units, for example.

Does anyone have an experience with that site? Is it a worthwhile tool?

[Bart_van_der_Wolf]

Does anyone have an experience with that site? Is it a worthwhile tool?

Hi Bob,

Since there are somewhat different methods of computing gamut volumes used by different applications, as long as you compare profiles on that website against eachother, you'll be fine.

Cheers,
Bart

[digitaldog]

Since there are somewhat different methods of computing gamut volumes used by different applications, as long as you compare profiles on that website against eachother, you'll be fine.

Which method is correct (does that matter)?

[Bart_van_der_Wolf]

Which method is correct (does that matter)?

Hi Andrew,

Correct is a big word, as they are all approximations based on some assumptions (e.g. which delta E method is used, or how is rounding done, and at which precision is the calculation done, to name a few). And no, it usually doesn't matter much. But comparing the volume (even of the same profile) from one application against that from another application, may/will show some differences in total volume. Hence my suggestion to stay within the same (whatever) application for comparisons.

Cheers,
Bart

[digitaldog]

Correct is a big word, as they are all approximations based on some assumptions (e.g. which delta E method is used, or how is rounding done, and at which precision is the calculation done, to name a few). And no, it usually doesn't matter much. But comparing the volume (even of the same profile) from one application against that from another application, may/will show some differences in total volume. Hence my suggestion to stay within the same (whatever) application for comparisons.

I'd totally agree to stick with apples to apples comparisons, but it would be useful to know which is closer to the goal and how the dE formula applies. We have differing dE formulas and there are reasons we'd use one over the other (dE 2000 is said to be 'more accurate' (better) with smaller dE values being calculated).

[Mark D Segal]

Well, you've shown a discrepancy of about 20% between PatchTool and ColorThink Pro, the latter with the calculation methodology stated. Some people may question a procedure starting with the dE (1976) formula vs the later ones. So it makes one wonder whether there is a "correct" method (based on what criteria? what purpose?) and whether the comparison of values within any one approach is perhaps more relevant and useful. Not sure - as it's hard to relate gamut volume to tangible reality in prints of real-world photographs, except when differences are truly large. I've considered it more prospective using these measurements - within one tool (CTP in my case) as indicatively useful comparative markers to predict how my image editing approach may differ between substantially different papers and what kind of limitations I should expect to encounter.

[Mark D Segal]

I'd totally agree to stick with apples to apples comparisons, but it would be useful to know which is closer to the goal and how the dE formula applies. We have differing dE formulas and there are reasons we'd use one over the other (dE 2000 is said to be 'more accurate' (better) with smaller dE values being calculated).

What goal?

If dE 2000 is "more accurate", would you consider CTP to be less useful relative to "a goal", or does it remain just as valid as any other approach if being used for comparative purposes?

I.O.W. how does one relate different "goals" to different "methodologies" in gamut volume calculations?

[digitaldog]

What goal?
If dE 2000 is "more accurate", would you consider CTP to be less useful relative to "a goal", or does it remain just as valid as any other approach if being used for comparative purposes?
I.O.W. how does one relate different "goals" to different "methodologies" in gamut volume calculations?

All excellent questions Mark. I can't answer them. Gamut Volume isn't something I've paid much attention to, not sure it's even useful 
For example, I have seen cases where the reported Gamut Volume for sRGB is a larger number than some output devices. But when I plot the two in 3D, I see where colors clip from sRGB should I use that working space. I think that's more useful. Or let's say more telling. Adding the ambiguity of how gamut volume might be calculated makes that metric even more confusing as to it's goal. At least on the Colorwiki for ColorThink, there's not much I can find about how this metric is useful: http://www.colorwiki.com/wiki/Evaluate_Device_Gamut_Volume

[howardm]

My personal take is it's not so much the final 'volume' # but where and how the volume is distributed ala 2 & 3D space rendering.

Case in point:  Hahn PR308 may have a greater volume than some other paper but 308 has deficiencies in the dark greens so
even if the volume is 'better' or 'greater', it may not be the optimal choice.

Feels like a lot of measurbating to me.

YMMV, no colors were harmed in the production of this posting. 

[digitaldog]

My personal take is it's not so much the final 'volume' # but where and how the volume is distributed ala 2 & 3D space rendering.

Exactly! It's like looking at some spec for the weight of a car alone, when you're wondering about it's gas mileage which we know vary. I don't see the number being very useful but I'd love to know why that might be an incorrect opinion (I'd love to learn more).

[Mark D Segal]

All excellent questions Mark. I can't answer them. Gamut Volume isn't something I've paid much attention to, not sure it's even useful 
For example, I have seen cases where the reported Gamut Volume for sRGB is a larger number than some output devices. But when I plot the two in 3D, I see where colors clip from sRGB should I use that working space. I think that's more useful. Or let's say more telling. Adding the ambiguity of how gamut volume might be calculated makes that metric even more confusing as to it's goal. At least on the Colorwiki for ColorThink, there's not much I can find about how this metric is useful: http://www.colorwiki.com/wiki/Evaluate_Device_Gamut_Volume

OK, we're largely in the same quandary. That said, I see this as being possibly useful at two levels, and this applies at least to CTP, which is the application I've been using. (1) Comparing gamut volumes between papers or printers does tell us something about the relative gamut limitations of our printing environment; this <could be> indicative of differing image editing challenges. (2) A bit geekier, one can use CTP to graph the gamut of a photographic image (reduced size so the program can cope) and compare it to the gamut of the printer/paper combination. This would provide an indication of whether the latter is likely to be constraining the appearance of the former. In both of these cases CTP is used for relative rather than absolute measurements, the purpose is conveyed by the nature of the comparison being made and the comparison is indicative rather than firmly predictive.

[Mark D Segal]

Case in point:  Hahn PR308 may have a greater volume than some other paper but 308 has deficiencies in the dark greens so
even if the volume is 'better' or 'greater', it may not be the optimal choice.

Yes I agree - the shape can be as important as the volume. Again, however, where the rubber should hit the road, is how to relate these numbers to observed real-world outcomes. That is where the CTP image graphing tool may be handy, but it would seem ones needs to do this often enough to make it useful for predictive purposes. Perhaps it's easier and faster just to run a few comparison prints with test images and we know all we need to know without the "measurebating". :-) ?

[Bart_van_der_Wolf]

My personal take is it's not so much the final 'volume' # but where and how the volume is distributed ala 2 & 3D space rendering.

Case in point:  Hahn PR308 may have a greater volume than some other paper but 308 has deficiencies in the dark greens so
even if the volume is 'better' or 'greater', it may not be the optimal choice.

Indeed, and that's why it's more important to compare the specific shape of the hul, and in particular for colors that are in the images we want to output (!)

The Argyll Colormanagement System allows to do that with these command line utilities that also produce a VRML file that can be viewed an appropriate browser plugin:
iccgamut      Create a gamut file or VRML file of the color gamut of an ICC profile.
tiffgamut      Create a gamut file or VRML file of the color gamut of a TIFF or JPEG image.
viewgam      Convert one or more gamuts into a VRML 3D visualization file. Compute an intersection.

Cheers,
Bart