RE: [Lcms-user] How good is the Profiler for slide film?

[Sullivan, Gregory (SNL)]

Title: Message

Marti, Thanks for the feedback. I’m glad the profile looks basically ok. But I don’t understand why the gamut is so enormous. If you superimpose even the sRGB trangle over my scanner profile, there is more green gamut in sRGB. And sRGB is considered a rather small gamut. Is it the case that slide film doesn’t have a very large green gamut?   I haven’t tried the improved profile you sent me yet, but thanks for building it.   Regards, Greg. p.s Still can’t get over the fact that this software is freeware!! J   -----Original Message----- From: Mart� Maria [mailto:[EMAIL PROTECTED]] Sent: Wednesday, 22 January 2003 10:41 PM To: [EMAIL PROTECTED] Subject: Re: [Lcms-user] How good is the Profiler for slide film?   Hi   I've taken a look on the profile. (found it on http://users.bigpond.net.au/sullivag/profiles/, thank you for making it available)   Seems quite good. The dE is really low, and after a close inspection, the gamut hull also looks reasonable.   There are some minor issues, like the white point not being perfectly on 255, and some mess on very dark shadows, but I believe the profile is absolutely useable.  I've tried to rebuild the profile with my current development sources, and these small problems are gone. I am emailing you privately a copy of this test profile.   Regarding primaries... well, scanners doesn't really have primaries, this is only a matter of information on what would be the theoretical primaries of colorspace. Anyway, the numbers you report makes the triangle (the gamut) enormous!   The fact green is outside viewing gamut does not have really any importance. You can interpret it as "which green should I capture to get 0, 255, 0?" Since the scanner would never give 0, 255, 0, the theoretical green is so saturated that is outside the visible spectrum, but otherwise this is not physically realizable. So don't worry about that. Also, the profile is LUT-based, and not using these numbers for the real work.   Regards, Mart� Maria The little cms project http://www.littlecms.com [EMAIL PROTECTED]     ----- Original Message ----- From: Sullivan, Gregory (SNL) To: [EMAIL PROTECTED] Sent: Wednesday, January 22, 2003 8:22 AM Subject: [Lcms-user] How good is the Profiler for slide film?   Hi, I've profiled my Nikon Coolscan IV ED using the LCMS profiling utilities, and Wolf Faust's provia/astia/sensia target. I think the profile is working well. However, a recent thread has started in newsgroup comp.periphs.scanners by Jeff Randall, who has noticed what appears to be a lack of green gamut in his Nikon LS4000 scanner, after he inspected the profile.  I've noticed the same characteristic in the profile I created, despite the fact that I haven't observed any noticable problems with greens in real scans. The person who started the thread says that he's used both LCMS (as I did), and also a beta profiling package, with essentially identical results from both.    The CIExy chromacities for my profile are: R:0.63, 0.35  G:0.37, 0.63  B:0.021, 4.4e-005 Are the blue coordinates of concern, being outside the visible spectrum? Does that suggest that the profile has a problem? Note that my coordinates are very close to Jeff's. (the LS4000 is very similar to the Coolscan IV/LS40)   The readme for Wolf's targets warns that not all profilers can handle the large gamut of slide film. So, how well has the LCMS profiler been tested with slide film?   Thanks, Greg.