Perceptual Rendering Argyll CMS

[DaveRichardson]

When producing a printer profile in Argyll I am asked to supply a profile that the picture will be converted from (eg ARGB)  in order that the perceptual rendering will be correct. Why is this, and if it is needed why did X-rites i1 match, the software I used to use, not ask for the same?

I ask this only for my own education, as it is not difficult to supply that input nor even build 3 separate profiles to convert from sRGB, ARGB and proPhoto.

Dave

[Doug Gray]

When producing a printer profile in Argyll I am asked to supply a profile that the picture will be converted from (eg ARGB)  in order that the perceptual rendering will be correct. Why is this, and if it is needed why did X-rites i1 match, the software I used to use, not ask for the same?

I ask this only for my own education, as it is not difficult to supply that input nor even build 3 separate profiles to convert from sRGB, ARGB and proPhoto.

Dave

One possible reason could be that they expand the PerCol gamut a bit to encompass what a printer can print given the limitations of sRGB. Since many, if not most, people that use PerCol, are doing so from an sRGB source it might make some sense. PerCol, is, after all, designed to look pretty and not map the source gamut as close as possible.  The problem, of course, is that printing processes don't intrinsically have any idea what the color sources are since all it sees is L*a*b*.

That's one of the reasons I dislike PerCol. You never know what assumptions the profile maker had.

[DaveRichardson]

Doug
Thanks for replying. A couple of things I don't understand - please excuse my ignorance.

First your reference to "the PerCol gamut". When I google "Percol" I get a return on paper emulsion (well that and coffee    !). What is PerCol?

Second why a limitation of sRGB comes into it. I am creating profiles on an Epson3800, measuring patches created from Argyll CMS and printed with no colour management with an i1pro spectrophotometer. The need to supply an input profile is described in the Argyll documentation as required only for the perceptual intent. I don't use sRGB at all - except when preparing images for the web.

There is a description on the Argyll CMS website but after reading it I have still not grasped the reason why 

Thanks

Dave

[digitaldog]

First your reference to "the PerCol gamut". When I google "Percol" I get a return on paper emulsion (well that and coffee    !). What is PerCol?

New term for me too! Probably Perceptual Rendering intent?

Second why a limitation of sRGB comes into it

By the time an output profile 'gets' it's data, it's in Lab or similar (from the PCS). Some assumption is can be made as to the gamut of the source. And AFAIK, this is true for all rendering intents! Now is it sRGB? I believe that depends on who's making the profile. So there are more than one set of assumptions here! 
Best to contact Graeme Gill and ask him specifically what's going on here! graeme@argyllcms.com

[DaveRichardson]

Thanks Andrew - makes sense. It could be that the X-Rite software was just making an assumption whereas the Argyll software is asking for reality, so that the printer profile perceptual rendering based on an sRGB source is "squeezed" less than that based on Pro-Photo source.


I'll drop Graeme a line.

Dave

[digitaldog]

It could be that the X-Rite software was just making an assumption whereas the Argyll software is asking for reality, so that the printer profile perceptual rendering based on an sRGB source is "squeezed" less than that based on Pro-Photo source.

Perhaps. This is supposed to be a major feature of V4 profiles, namely what's called PRMG. X-rite doesn't support it. So their V4 profiles are just V2 in sheep's clothing. But what Argyll is doing is just a guess, never used it. But Graeme is very open about what's going on and can answer your questions without ambiguities. Plus he does hang out here and posts regularly so you might just try sending him a private message.

[GWGill]

When producing a printer profile in Argyll I am asked to supply a profile that the picture will be converted from (eg ARGB)  in order that the perceptual rendering will be correct. Why is this, and if it is needed why did X-rites i1 match, the software I used to use, not ask for the same?

Some information about this is here.

Short answer - to be able to do actual gamut mapping, rather than non-specific compression or clipping.

Longer answer - to be able to precisely map one gamut to fit inside another, you need to know what the gamuts are. When you are making a display or output profile, you know what one of the gamuts is (the output), but you know nothing specific about the other. Hence the need to specify it.

I have little idea what i1 match does. Since you tell me that no source gamut is able to be provided, I can only guess that it assumes a source gamut, or does a general level of compression or clipping of an incoming color that gets close to the output gamut.

An idea that has been floated by the ICC in relation to ICCV4, and (implicitly) may be something that some profile makers have been doing for a long time, is to target some common, intermediate gamut, making use of gamut mapping in both the A2B and B2A tables. The above link covers the drawbacks of this approach. In summary - you have still lost the relationship between source and destination gamut, leaving the overall gamut mapping in the dark about what it should do, and you end up concatenating two gamut mappings, increasing the inaccuracy of the whole process.

[Doug Gray]

All of which, with the addition of PRMG confusion, is why Perceptual Colorimetric rendering, is something I avoid. There simply is no way to guarantee that you can print something with different printers/profiles and know you will get consistent results. You might. You just don't know. For RelCol you will always get consistent results printing in gamut colors. At least if your profile is properly made.

GWGill's link explains the conundrum pretty well.

[GWGill]

All of which, with the addition of PRMG confusion, is why Perceptual Colorimetric rendering, is something I avoid. There simply is no way to guarantee that you can print something with different printers/profiles and know you will get consistent results.

The whole point of how ArgyllCMS does gamut mapping, is that you know exactly what is going on. Other profile makers, not so much...

[Doug Gray]

The whole point of how ArgyllCMS does gamut mapping, is that you know exactly what is going on. Other profile makes, not so much...

How about not at all! 

Any idea how Argyll's mapping of out of gamut colors is done compared to other profile vendors in BtoA1? I've been looking at the difference between conversions of extreme ProPhoto RGB values going directly to the PCS v clipping to Adobe RGB first. The printer gamut boundaries are inside Adobe RGB much of the time and the conversion path using Adobe RGB as an intermediate seems to produce significantly larger dEs on average.

[DaveRichardson]

Short answer - to be able to do actual gamut mapping, rather than non-specific compression or clipping.

Longer answer - to be able to precisely map one gamut to fit inside another, you need to know what the gamuts are. When you are making a display or output profile, you know what one of the gamuts is (the output), but you know nothing specific about the other. Hence the need to specify it.

Graeme - thanks , that answers my question perfectly, I had read the article in your link but your addition here has clarified it for me. I can see the difference in compression if I apply, to the same image, a profile based on conversion from sRGB to another based on conversion from ProPhoto, both with perceptual rendering.

Perceptual Colorimetric rendering

Doug - thanks for clarifying.

99% of the time I use the Relative Colorimetric intent - but there is always the odd image with colours where that does not work was well as the Perceptual intent (some flower images come to mind).

Dave

[digitaldog]

Perceptual Colorimetric rendering

Is that a new or made up term?

[digitaldog]

How about not at all! 

Your bias and prejudice that Perceptual rendering isn't appropriate for you has been made clear more than once. The idea that there is no merit in any perceptual renderings from all products for all users is to be kind, hogwash. Again, profiles know nothing about color in context. I asked you in another thread why photographers shouldn't be given a choice in color rendering in E6 file (Velvia vs. Ektachrome as an example) which is no different than a Perceptual mapping of colors, you ignored that question. IF you need a colorimetric reproduction, Perceptual isn't for you. Yet many image makers desire pleasing color and that's exactly what a Perceptual rendering attempts, repeat, attempts to provide and not all are created equally. If you're quest is to argue with the author of ArgyllCMS he shouldn't be providing a Perceptual rendering in his product, I'm going to sit back and really enjoy that debate. 

[digitaldog]

Other profile makers, not so much...

Or not at all, at least from X-rite. You have controls to alter the perceptual rendering attributes but they make no sense and the scales used (usually a 0-100 slider) makes absolutely no sense. You're flying blind.

[DaveRichardson]

You have controls to alter the perceptual rendering attributes but they make no sense and the scales used (usually a 0-100 slider) makes absolutely no sense. You're flying blind

On the i1match software - there was not even a slider, although I did use a "hack" given by Eric Chan a few years ago that allowed 3 variations on the perceptual intent.
I am enjoying getting to grips with the Argyll CMS software and the control that it allows. It is like switching off "programmed -auto" on the camera and taking control.

Dave

[DaveRichardson]

As always one answer leads to another question.

My current workflow uses Adobe Bridge, ACR and Photoshop CS6. However I will shortly be moving to Lightroom which I understand uses a variation on Prophoto colour space but with a gamma of 1.0. Is that colorspace supplied with the software as an icc/icm file that I can use as the source for building custom printer profiles to print direct from light room?

Dave

[bjanes]

When producing a printer profile in Argyll I am asked to supply a profile that the picture will be converted from (eg ARGB)  in order that the perceptual rendering will be correct. Why is this, and if it is needed why did X-rites i1 match, the software I used to use, not ask for the same?

Some information about this is here.

Short answer - to be able to do actual gamut mapping, rather than non-specific compression or clipping.

Longer answer - to be able to precisely map one gamut to fit inside another, you need to know what the gamuts are. When you are making a display or output profile, you know what one of the gamuts is (the output), but you know nothing specific about the other. Hence the need to specify it.

When one supplies the requested picture, does the mapping software examine the gamut of colors within the image or the gamut of the color space (container for the image)? If the latter, one would not really need a picture, but merely a reference to the color space.

Thanks,

Bill

[Doug Gray]

Is that a new or made up term?

It's made up in the sense that I use it to refer to perceptual intent rendering that is reversible or has deterministic colorimetric rendering. For instance, if I like the perceptual rendering and want to be able to reproduce that at some future time or using a different printer I'll to an AtoB using RC and save the rendered image. Then, just a simple RC print will make the same highly consistent perceptually rendered prints. Works great unless you are printing to a smaller gamut like going from glossy to matte.

[Doug Gray]

As always one answer leads to another question.

My current workflow uses Adobe Bridge, ACR and Photoshop CS6. However I will shortly be moving to Lightroom which I understand uses a variation on Prophoto colour space but with a gamma of 1.0. Is that colorspace supplied with the software as an icc/icm file that I can use as the source for building custom printer profiles to print direct from light room?

Dave

No need. All working space gammas are converted to gamma=1 before they go to the PCS profiles use. I made a gamma=1 version of Adobe RGB(1998) and ProPhoto RGB and use it when I want to look at the RGB channels in a linear luminance sense. Mostly I use that when checking gray scale reflectance accuracy. Useful quick checking for lens glare or accidental conversion when one wanted scene referenced conversion. I.e., linear v S-curve.  Not something I often use but can sometimes be handy. If I'm looking at an 18% gray I'd like it to read 18% instead of somewhere in the middle of the working range. I can't think of a use in normal situations.

[digitaldog]

On the i1match software - there was not even a slider, although I did use a "hack" given by Eric Chan a few years ago that allowed 3 variations on the perceptual intent.

Sorry, I wasn't totally clear. The settings I referred to were from i1Profiler.