Hi Elle,
You are right, gamut check was not intended to work in unbounded mode and also, due to the way it works, it needs a space perceptually uniform to perform good. This is because gamut check uses a LUT driven by CIE L*a*b* to hold the distance of each color to gamut boundary. Works well for traditional 8-bit imaging but definitevely not for linear spaces close to XYZ. This is not a bug, but a limitation due to the algorithm used. I take two action items: fully de-activate gamut check on unbounded mode and document the limitation. Will happen in 2.9 Thanks for reporting Marti. Elle Stone <ellest...@ninedegreesbelow.com> escribió: > On 08/10/2016 06:27 AM, Elle Stone wrote: >> Hi Marti and all, >> >> When soft proofing with high bit depth GIMP, which uses LCMS, the gamut >> check isn't reliable. I've confirmed some of the issues using tificc at >> the command line, so it's not GIMP code per se. >> >> 1. When using the gamut check, when soft proofing to a destination color >> space that supports unbounded ICC profile conversions, the "bounded" >> color gamut of the source color space affects results. So colors that >> are marked as out of gamut in one source color space will show as in >> gamut in another source color space. In general, the closer the source >> and destination color space gamuts are, the larger the proportion of >> colors that are shown as in gamut with respect to the destination color >> space. This is true even if the entire image is actually out of gamut >> with respect to the selected destination color space. >> >> 2. The gamut check results vary depending on the TRC - this affects soft >> proofing not just to profiles that support unbounded conversions, but >> also to LUT printer profiles that don't support unbounded conversions: >> >> * The LCMS gamut check gives different results depending on >> whether the source color space has a linear gamma TRC or else has a more >> perceptually uniform TRC >> >> * When editing in a linear gamma RGB color space, the LCMS gamut >> check can easily show most or all of the image as out of gamut, even >> when the entire color gamut of the source image does fit in the >> destination color space's color gamut. Dark and midtone colors are >> marked as out of gamut. >> >> * For source color spaces with more or less perceptually uniform >> TRCs it seems the exact TRC of the source color space makes a difference >> in the results of soft proofing. >> >> Best regards, >> Elle >> > > In case it might be helpful, here are links with screenshots > illustrating some problems with LCMS gamut checks: > > * http://ninedegreesbelow.com/files/soft-proof/soft-proofing.html (this > link has sample images if anyone wants to verify some of the issues with > current LCMS soft proofing, and also screenshots comparing soft proofing > results from several different image editors). > > * http://ninedegreesbelow.com/bug-reports/soft-proofing-problems.html > (screenshots just from GIMP 2.9) > > Currently it seems that LCMS gamut checks only give correct results if > several conditions are met: > > 1. The source color profile's "bounded/clipped/display range" color > gamut must encompass all the image colors. > This means means soft proofing while editing using > unbounded/unclipped floating point processing only gives correct results > if in fact no image colors are out of gamut with respect to the color > space in which the image is being edited. > > 2. The source color profile must have a perceptually uniform TRC. > This means anyone editing in a linear gamma color space sees > incorrect gamut checks. > > Also the destination color space can't support "unbounded/unclipped" > profile conversions, but this topic is already covered in other threads. > > Also currently LCMS gamut checks don't take into account whether or not > "black point compensation" has been selected. I can see arguments either > way, but apparently PhotoShop does produce different gamut checks > depending on whether BPC is activated or not. > > Best regards, > Elle > -- > http://ninedegreesbelow.com > Color management and free/libre photography > > ------------------------------------------------------------------------------ > Developer Access Program for Intel Xeon Phi Processors > Access to Intel Xeon Phi processor-based developer platforms. > With one year of Intel Parallel Studio XE. > Training and support from Colfax. > Order your platform today. http://sdm.link/xeonphi > _______________________________________________ > Lcms-user mailing list > Lcms-user@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/lcms-user ------------------------------------------------------------------------------ Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi _______________________________________________ Lcms-user mailing list Lcms-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lcms-user
