Media Players for PC monitors calibrated using .icm profiles

[sanlyn]

Anyone know of a media player (MPEG, AVI, M2v is all I need) that recognizes or can be set to recognize WinXP .icm monitor profiles? I just spent 2 days downloading every player in sight, and the best I found has unsatisfactory Tint, Contrast, etc., which don't even get into the ballpark in trying to mimick my monitor's calibrated settings.

The proper way to calibrate a PC monitor is with colorimeters and software from iMatch/EyeOne or Spyder. This stuff works really well, but they adjust your monitor via LUT (monitor driver lookup tables) that are stored in .icm color profiles. The monitor's menu controls are set at factory defaults (yecch!) to allow these calibrations to work. Some software apps can be directed to recognize your color management system, such as .icm's (Photoshop, a few editing and photo programs, etc.). As far as I can tell, VirtualDub and many VDub filter preview windows seem to be automatically keying-in to the .icm profile on my PC's. So does TMPGenc Plus 2.5. TMPGenc Editor v2 apparently uses it if you tell its Preferences to ignore overlays, but the effect is visible only in the Preview window, not in the cut/edit windows. Other stuff wouldn't know an .icm oftware defrom a .txt (TMPGenc Author hasn't a clue, neither does Adobe's lesser graphics apps and freebies).

Surely some clever graphics software designer (ahem) has come up with a media player that recognizes i.cm color management profiles and doesn't keep using those stupid monitor presets.

Yes, I know about setting graphic card overlays, etc., but ya see folks, that doesn't even come close to a proper calibration, and it undoes the work of the .icm profile to begin with.

[jagabo]

I believe the ICM is applied at the graphics card's DAC output (or, for digital monitors the adjustments are made in the output driver) so any program that writes RGB to the Windows GDI will automatically get corrected colors.

Try using VLC with output set to Windows GDI. Your levels will be wrong though. I don't see any way to get VLC to use a rec.601/709 matrix instead of pc.601/709 matrix to convert video YUV to RGB.

[sanlyn]

Thanks, jagabo, that might lead somewhere. However, the calibration is for sRGB. Has nothing to do with Rec.601/709, etc. Those "levels" I'm not worried about, those are checked with other tools. etc, and $5000 pro monitors have such setup options, but not a $250 Samsung.

Now, I suspect my Windows desktop is somehow hooking into that ICM. so possibly it's getting there via the path you suggest ? ? The "wallpaper" is a ten-step 352x240 grayscale in center-screen, and the background wallpaper is RGB-32 gray. At startup these grays look mostly cyan, with reddish darks. When the loaded ICM hooks-in a few seconds later, the grays snap-to; they look perfectly neutral...well, OK, RGB16 is a tad red, but you have to look for it on this $200 LCD. So my ultimate test to check if display software is using the ICM profile is to play or display a monochrome source. Without the ICM effects, b&w looks cyanish. With the ICM in play, it looks neutral.

Most Windows dialogs must be using the ICM, too, because without the ICM loaded many of the gray panels look cyanish and my medium gray MS-Word background has the same tint. With the ICM loaded, colors appear to be correct (However, most of the dialog-window panels and scroll bars aren't really "gray". By default they're R-212 / G-208 / B-200. Which is to say they're not neutral gray, they're slightly yellow. So if these things don't look at least a little gray, somwthing's wrong!)

I'll try it in VLC. But I'm shocked, shocked to find no ICM functionality in these media players.

[sanlyn]

I believe the ICM is applied at the graphics card's DAC output (or, for digital monitors the adjustments are made in the output driver)

Adjustments are made via the monitor driver's LUT (LookUpTable values), which are adjusted and loaded into the new, default .icm profile. The new profile's name is not the same name as the monitor's supplied .icm (driver) file. However, I've seen warnings that some graphic cards' control panel overlay settings monopolize system color management and override all .icm's. Calibration software had the problem of being blocked by the ATI Control Panel in Catalyst versions earlier than Cat 3.9. A $15,000 pro display card comes with a calibration report -- for the card only, not for what it's hooked into. In effect, running calibration software/hardware on a PC is akin to adjusting your monitor to your graphics card's output. Change graphics cards, and you have to recalibrate your monitor with a new .icm profile -- even if the new card is a replacement copy of the old one.

so any program that writes RGB to the Windows GDI will automatically get corrected colors.

Doesn't appear to be happening that way. If it worked that automatically and always, why do Adobe and other hifalootin' graphics apps display so poorly until you point them to a custom .icm? I think these players default to the Windows Generic sRGB .icm and ignore all else. MediaPlayer and PowerDVD are prime examples, which is why I can't use them to "test" videos, but have to trust VDub and TMPGenc - which are serving me well with color correcting, thanks to their viewing screens, but they're mighty tiny windows. It's also why I won't watch movies on a PC. Can't fathom why so many buyers want 50-yard widescren PC monitors to watch lurid bdsm cartoons thru funky software like PowerDVD in "Vivid" mode.

Try using VLC with output set to Windows GDI.

Unable to find a setting that looks like that in VLC.

Your levels will be wrong though. I don't see any way to get VLC to use a rec.601/709 matrix instead of pc.601/709 matrix to convert video YUV to RGB.

Not trying to get VLC to make matrix conversions. Just looking for a media player that's useful with monitor calibrations that don't depend on those godawful-useless LCD menu settings.

If nothing's available maybe I should learn Java and modify VLC to use monitor .icm's the way Photoshop does do. But that would take a long, long time.

But wait -- Clue #1: I'm told that media players don't use .icm profiles. They use default monitor settings, or whatever settings owners have set in their totally-useless-but looks-hitech user menus. Those menu settings aren't stored in the registry (how would they get there from your monitor?) The LUT values originate in the monitor's .icm file (which is why you get poor results using a Spyder2 to calibrate a monitor with only the generic plug-n-play-monitor .icm for a driver). So these players are using a "default" that comes from somewehre, but where?

Clue #2: TMPGenc MPEG Editor v2 has a Preferences setting for renderers. There are 3 choices. I don't remember all 3 right now, but the top-listed choice is the default in every new TMPGenc project, and uses some layer or other. The middle choice is "Do Not Use Overlay". If you chek "Don't use" and save it, you'll clearly see the color balance change in the preview screen (but not in the cut/edit window. Why?). I've never used those godawful-useless-anyway layer options in any graphics card anywhere, so why does the color balance change when you disable an overlay that was never set anyway? There really are creatures on this planet that are called overlay-enabled or overlay-aware (if you require a more P.C. term) mediaplayers.

Looks like another long weekend of research on the 'net. Good thing it's Labor Day.

[jagabo]

Media players normally use Video Overlay. Video goes straight to the graphics memory in YUV form (usually YUY2 or YV12) and the graphics card has a separate DAC that converts to RGB for the monitor (which apparently doesn't use the main DAC's ICM LUT.

In VLC: Tools -> Preferences -> Video (left pane) -> Output (right pane) -> Windows GDI Video Output. If you use this option VLC will convert from YUV to RGB before writing the image to the screen with Windows GDI. Unfortunately I don't see any control over what matrix to use. It appears to use the PC.601 matrix (same matrix as AviSynth's ConvertToRGB(matrix="PC.601"). This does not perform the normal luma expansion for displaying video on a computer monitor.

I believe some apps have their own ICM handling because Windows didn't have this functionality many years ago. Or one might want to use a different ICM while editing and not have to change the system wide ICM.

[sanlyn]

Thanks for all this info, jagabo, you're a scholar and a gentleman. The wife controls all activity on weekends, but I'll look for those settings soon (must be getting too old for this, I looked everywhere). As for the matrix, PC.601 is fine and at least consistent with everything else. Can always use VDub or AviSynth for that -- short of paying 4 figures-plus for a pro monitor.

Guess I have some programming update work ahead if I'm going to figure out the grimy details underneath all this ICM technology. Thanks again.

[jagabo]

It looks like I was wrong about how the display properties ICM profiles are used (ie, what you get with Display Properties -> Settings -> Advanced -> Color Management). It does not load a system wide default profile. All it does is offer a suggestion to ICM aware applications. So indeed, programs must be designed specifically to use ICM.

But there is some way of getting system wide color changes. Of course, there are the color controls in in the graphics card's setup applet. But those are usually pretty crude (contrast, brightness, color, hue and gamma sliders) so it would be hard to get a perfect adjustment. And video overlay has separate settings. I found a program called QuickGamma and it effects the entire system (except video overlay). When using VLC set to Windows GDI output the video was effected just like everything else on the desktop.

QuickGamma only has crude controls (overall, red, green, and blue gamma sliders) so you can't get precise adjustments with it. But the implication is there should be some way to get more precise color adjustments that are system wide. Except, again, video overlay. But many media players can also be set up for VMR9 output (I believe this is the default video output device in Vista) which is effected by the QuickGamma settings.

http://quickgamma.de/infoen.html

[sanlyn]

Most helpful, jagabo. I formerly used QuickGamma. It worked OK with overall black/white levels. Back then I used a Sony Trinitron and AdobeGamma. I guess the Trinitron spoiled me; its factory settings were close to ideal and needed very little meddling.

Unfortunaely QUickGama undoes much of the work of calibration packages. One can get tricky with these calibrators; one can initially make QuickGama and monitor menu settings that improve overall imaging, and leave them in place so that the calibration software "works around" them in auto mode. The software itself doesn't know how you've set your monitor anyway; it knows only what it reads from its own test patterns, then makes LUT adjustments for a new .icm accordingly. Or you can use monitor defaults before running the calibration. But each display app reacts differently to each method. Either way, you're forced to carefully observe how display apps behave, and work from there.

This a.m. I'm playing with VLC to get somewhat closer to target. This is better than no adjustment at all or setting overlays that make everything go haywire. After I get all my personal tantrums out of my system I could probably live with it.

This leaves several considerations. (a) Some apps can be made to exploit ICM's -- e.g, Photoshop and the like. (b) A few apps use ICM's automatically but can't be told to un-use them, like some preview dialogs in VirtualDub filters, TMPGenc 2.5's image adjustment window, and--apparently--the Windows desktop wallpaper display. (c) Some apps "see" ICM's and show color OK but can't adjust to an ICM's luma levels. VirtualDub's default front-end viewer seems to work this way, but oddly its capture viewer seems to be reading ICM's more closely. I'll recheck that later, if I can figure how to do it reliably. (d) Then you have apps that can disable video overlays and thus might "see" ICM's but can't adjust overall luma, like TMPGenc MPEG Editor v2's preview screen (e) Then you have apps whose renderers use what amounts to anyone's guess. To me, all media players seem to take this route.

Guess I'll continue my search. Fiddling with open-source code and getting back into programming will likely keep me busy for some time. Meanwhile, picky fusspot geezers like me will have to make-do with what we have. Most of us are used to that by now. Or should be.

[jagabo]

I was thinking you could get QuickGamma, or something like it, to load an ICM as a global windows color correction. Then turn off ICM processing in all the software. Then only video overlay would have the wrong colors.

QuickGamma stores its lookup tables in [HKEY_CURRENT_USER\Software\Eberhard Werle\QuickGamma]. It shouldn't be too hard to extract the LUT from an ICC/ICM profile and write it to the Registry.

[sanlyn]

I was thinking you could get QuickGamma, or something like it, to load an ICM as a global windows color correction. Then turn off ICM processing in all the software. Then only video overlay would have the wrong colors.

That's a workable idea. My understanding is that calibration packages make RGB grayscale adjustments that affect gamma at certain extremes (you'd have to make the same adjustments if you were calibrating a tv, which is an entirely manual process). Gamma is the perceived relationship between NTSC IRE-10 levels and NTSC IRE-100 (for a PC that's about RGB-24 and RGB-235 or so). It's probably the easiest thing to adjust on any monitor; many test patches exist, or you can make your own. It's a recursive process. Since setting brightness (bleck level) and setting contrast (white level) affect each other, you go back and forth between light and dark patches until you get the best adjustment between blacks that aren't crushed and brights that aren't burned off. On a monitor or TV, if you have a colorimeter or other hardware to measure, you end up with black RGB-16 luminance that's .0065 times your RGB-235 luminance. Extremes are left to fall where they may, since they're often imperceptible anyway. Those levels supposedly give you a gamma of 2.2. I believe you mentioned you've used a Spyder2; that package will get you mighty close to this figure, DisplayOne will get ever slightly closer, but either will get close enough.

However -- oh, it's always so damn complicated -- each RGB color has its own gamma value, which varies from dark to bright. Much more complex arrangement. Calibrations affect the RGB grayscale and individual color luminance (saturation) as well. From what I gather, there's no "real", single gamma number: it's a curve, not a number. So grayscale and saturation adjustments affect that curve. If you run your Spyder2 program and watch it, you'll see it repeat a test-pattern sequence several times. Each pass is a reiteration that looks for the best compromise between grayscale color balance, color temp, and overall gamma across the spectrum.

Add to that: some monitors have what they like to call "gamma" settings. Samsungs and LG's have "Mode" choices. While each choice looks as if it changes bright/contrast, you'll clearly see that the color balance changes as well, especially in darks and brights.

Your point is a good one, though. The question is how to turn off ICM in software that doesn't allow it (the registry perjhaps? Shudder!).

Meanwhile, after an hour using your suggestions with VLC I've managed to reach a compromise. Color's now a bit closer to the ICM apps, bright/contrast gets even closer to target. Anyway, the final test uses 2 calibrated tv's and an old calibrated Hitachi CRT (that's a 25-minute drive to my father-in-law's). As for other media players, I give up. PowerDVD has an interesting 2-part color control instead of a single Tint or Hue, but it's just too tricky to work with.

Now, if I can have another 18 months to learn Java and see what I can do with VLC's code base . . .

[jagabo]

QuickGamma stores three LUTs in the Registry. One for each color primary. These are just lists of 256 16-bit integers. You can use any curves you want (although QuickGamma only produces simple gamma curves). A quick hex dump of some ICM files seemed to show similar lists (along with some meta data).

If you can't turn off ICM in a program just set it to a linear ICM.

Most media players will allow you to disable video overlay. Or just open one video in a player with video overlay (don't minimize it) before opening the program you want to use. Since only one program can use video overlay the second will use Windows GDI.

[sanlyn]

I disabled overlays and set some image adjustments, but VLC is the only setup (so far) that seems satisfactory. BTW, image adjustments I make with VirtualDub tools and TMPGenc seem just about spot-on when I play the videos on tv or view captures in Photoshop. Seldom need a re-do. That's a huge improvement over what I had before I used the Spyder or EyeOne.

Registry: that would keep everyone busy 'til the apocalypse. I tried the same searches a whle back. I just noticed something: my old Spyder2 saved .icm files, but my EyeOne saves ".icc" files. I'm on my laptop now, which involved adjustments simpler than a desktop LCD. An ICC apparently has the same binary info as ICM, but in ICC it's preceded by dozens of lines of decimal tabular values and the number of reiterations made (42 in this case). Will check my desktop for this after Labor Day when my wife frees my 'puter.

A sample of the last 4 decimal lines:

40	G4	11.07	 6.60	1.79	30.88	40.98	25.06
41	G5	16.89	 9.95	2.34	37.76	48.03	31.72
42	G6	23.64	13.83	2.77	43.99	54.35	38.87
END_DATA

All have the identical tabular format. Must be over 100 in all. No column headings. At the end are some English-language result stats (Say, I managed to get a gamma of 2.0 on this cheap laptop. Not bad. Best luminance level was 78--a bit below the ideal 120, but this is a $550 laptop for goodnesssake). At the end is the usual unprintable ICM gobbledygook.

Decided to install VisualStudio 6 on my PC's. I re-installed or built new PC's when I retired, but omited VS-6 (just to keep me from driving myself nuts). Restoring VS-6 gives access to C++ multimedia and hardware SDKs. I don't know what language creates ICMs or ICCs. Guess I can't resist getting back into the fray. Downloaded Java's kit, too.

[vhelp]

Hi sanlyn..

Knowing where the data is actaully kept would help. We would then only need to overlay on top of the old values..save as a new icm template w/ unique name for each type of project we need for given work in color levels.

The registry idea..I'm on windows 98 so maybe its limited to what can be copy/paste-able to a notepad. win98 did not allow me to copy/paste, so this is crude.

I was thinking of just using one of these as a template and just overlay new values by writing them out to the registry. Sorry for going the "pic" route but I did explane why already

If my math was correct, I calculated 760 values for each template of .ICM file, in this settup:
HEX 0000-02f8 = 760 DEC, I think.
I won't post all pics here because it would take 7.1 snap shots and that would be a waist, so here's just a few of what the registery holds:






* 1. This is the one that starts off the DATA section in the first pic, at HEX position 0110
* 2. The other junk before and after must be header and other info.
* 3. Poking the values at hex 0110 by hand would not be effecient, but you could code a script or utility to do this.



* I'll take down the pics later so not to bog this discussion w/ too many images.

sanlyn, assuming that these are .icm templates (looking at the first pic) you could even test this out yourself, if you have a set of values you want to test, post them up and run the utility poke the values in, or by hand..there's only a hundred you said, right ? That should get you started.

To test, start up TMPG or whatever you said you used, and note the levels (if possible) and then run the poker utility with your new values and then re-run TMPG again and see if your colors change.

At least we now have a place to start, I think

-vhelp 5186

[jagabo]

Note you can export from the Registry by clicking LUT (in your example) and selecting Export. That will give you a text file that you can edit with Notepad (or whatever). The filename will end in .REG. If you double click on it the data will get loaded back into the registry (after a confirmation dialog).

A sample Registry export from QuickGamma:

Windows Registry Editor Version 5.00

[HKEY_CURRENT_USER\Software\Eberhard Werle\QuickGamma]
"RampData"=hex:00,00,01,01,02,02,03,03,04,04,05,05,06,06,07,07,\
  08,08,09,09,0a,0a,0b,0b,0c,0c,0d,0d,0e,0e,0f,0f,\
  10,10,11,11,12,12,13,13,14,14,15,15,16,16,17,17,\
(many lines removed)
  f0,f0,f1,f1,f2,f2,f3,f3,f4,f4,f5,f5,f6,f6,f7,f7,\
  f8,f8,f9,f9,fa,fa,fb,fb,fc,fc,fd,fd,fe,fe,ff,ff

There are 1536 values. Three tables of 256 16-bit integers (each int is saved as two bytes, little endian). The red table first, then green, then blue. These are simple lookup tables for converting RGB primaries. For example:

output_red = red_table[input_red] / 256;

Where output_red is the red value you see on the screen (ie, what is sent to the monitor), input_red is the red value stored in video memory (whatever a program wrote there), and red_table is the red array in the registry (obviously, this table is loaded into the graphics card's LUT by QuickGammaLoader and the graphics card performs the lookup as it displays the video).

[sanlyn]

At least we now have a place to start, I think

Inded, vhelp! Brilliant research. I was persuing "LUT" strings myself, but you beat me to it. Your tip led me to this page at Microsoft:
http://www.microsoft.com/downloads/details.aspx?familyid=1E33DCA0-7721-43CA-9174-7F8D4...displaylang=en

Win98 uses profiles differently -- which likely is why many calibration packages or software don't work with 98/Me (I don't think the free HCFR software works with 98. Requires Win2K or later). Photoshop-6 changed the way it loads color profiles starting with Win2K. I discoverd this by browsing DryCreekPhoto. Microsoft is again misusing Vista to push a new color management system so that nothing pre-Vista will work. Don't they ever learn?

In any case, this opens the possibility that whatever profile is being used by non-aware apps (probably sRGB) could have its values revised. Question: why would this LUT be in the registry when it already exists in .icm's/.ici's. BTW, the 100 or so decimal-coded lines in the .ici appear to be a record of trial readings. The actual LUT is in binary at the bottom of the file.

I think this thread has expanded into the area of programming, so if I post more I'll get it into the Programming forum. Meanwhile, my wife is demanding attanetion (I broke my promise and worked on this all day! ). So we're headed out for hot dogs and fries to carry to the beach so we can watch the sunset. It's cloudy now, but we'll watch the sunset anyway.

Thanks again, vhelp . Will delve further (farther?).

[vhelp]

Ahh, thank you for the registry tip, jagabo. Yeah, the registry is the devil. I have no place in my heart for such worship. Anyways., personally, I much prefer the old clunky .ini file route, but writing binary data like the LUT's is a good example of why M$ created the evil in the first place I guess.

. . .

-- I just got an idea while reviewing the registry trick by jagabo (thank yuu) the thought came to me. I'll finish reading the rest of the posts after I post.

As an excercise, why not use an image (maybe a reference one) of something you could use to gauge and then vary the icm LUT values and see how it effects things. I don't know if you have to be in overlay or just plain RGB or windows os platforms.

For instance, lets say we write a front-end or gui utility that does following:
1. open an image, and using the toa
2. features a couple of sliders to vary color R/G/B values individually
3. addes *new* calculated LUT values
4. save them in a .reg file, like in jagabo's example above.
6. re-run the utility and review for any changes.

Maybe that is too crude and makes no sense, but it seems to in my head at the moment. Its sunday, and I just came back from a 10 lap walk at the park.

I don't get much of these adrenlyn rushes like I used to, they are very rare indeed these days..used to be alot in the old days..so take advantage for now cause later on i'll prob go back into hybernation mode/

-vhelp 5187

[vhelp]

Question: why would this LUT be in the registry when it already exists in .icm's/.ici's.

maybe because it was part of a separate utility to change gamma I was using it a short while when I first got my new (first) widescreen monitor and learnt that something had changed. Then I started cursing out loud here for help (you must have been sleeping or something) and I went on a rampage to figure it all out..still figuring it all out. But, you discussion here is the only one that seems closest to anyone finally giving an ear to the trueth and problem. Personally, I feel its all about gamma and the issues (calibration) is governed by this aspect. Fix gamma and everything else falls into place, or should with a few other adjustments.

regarding your last comment .. M$ is evil.

-vhelp 5188

[vhelp]

EDITED: I had time to revise my latest comments..

I see that jababo posted a short snipplet and it seems to look similar to mine. I'm wondering if the LUT's setup is the same across all utility type apps that adjust gamma. Let me explain..

If we post whatever our registry has for LUT values maybe they follow a similar format, the binnary structure may be slighlity different, in size too, but the LUT values may follow the same location within the data block. Below is the registry setting for my gamma tools LUT values.

If the Start and End locations in the data block are the same, we can update those values correctly. But bare in mind that some LUT registry setup may be different in size and pertain to the gamma tool that created and make use of it, so that is something to expect in determining actual LUT locations in the registry.

If it helps narrow down the location, here's my registries LUT values.

* Using jagabo's registry method: menu, \Registry\Export Registry file\save as "LUT"
(or maybe there's a better way )

"Theater 2.2 Gamma D50"=hex:
  0e,0f,10,11,12,12,13,14,15,16,17,17,18,19,1a,1b,1c,\
  1d,1d,1e,1f,20,21,22,23,24,24,25,26,27,28,29,2a,2b,2c,2c,2d,2e,2f,30,31,32,\
  33,34,35,36,36,37,38,39,3a,3b,3c,3d,3e,3f,40,40,41,42,43,44,45,46,47,48,49,\
  4a,4b,4c,4c,4d,4e,4f,50,51,52,53,54,55,56,57,58,59,5a,5a,5b,5c,5d,5e,5f,60,\
  61,62,63,64,65,66,67,68,69,6a,6a,6b,6c,6d,6e,6f,70,71,72,73,74,75,76,77,78,\
  79,7a,7b,7c,7c,7d,7e,7f,80,81,82,83,84,85,86,87,88,89,8a,8b,8c,8d,8e,8f,90,\
  91,91,92,93,94,95,96,97,98,99,9a,9b,9c,9d,9e,9f,a0,a1,a2,a3,a4,a5,a6,a7,a8,\
  a9,aa,ab,ab,ac,ad,ae,af,b0,b1,b2,b3,b4,b5,b6,b7,b8,b9,ba,bb,bc,bd,be,bf,c0,\
  c1,c2,c3,c4,c5,c6,c7,c8,c9,ca,cb,cc,cd,cd,ce,cf,d0,d1,d2,d3,d4,d5,d6,d7,d8,\
  d9,da,db,dc,dd,de,df,e0,e1,e2,e3,e4,e5,e6,e7,e8,e9,ea,eb,ec,ed,ee,ef,f0,f1,\
  f2,f3,f4,f5,f6,f7,f8,f9,fa,fb,fc,fd,fe,ff,03,04,04,05,06,06,07,08,08,09,0a,\
  0a,0b,0c,0c,0d,0e,0f,0f,10,11,11,12,13,14,14,15,16,17,18,18,19,1a,1b,1b,1c,\
  1d,1e,1f,1f,20,21,22,23,23,24,25,26,27,27,28,29,2a,2b,2c,2c,2d,2e,2f,30,31,\
  31,32,33,34,35,36,36,37,38,39,3a,3b,3b,3c,3d,3e,3f,40,41,41,42,43,44,45,46,\
  47,48,48,49,4a,4b,4c,4d,4e,4f,4f,50,51,52,53,54,55,56,56,57,58,59,5a,5b,5c,\
  5d,5e,5e,5f,60,61,62,63,64,65,66,67,67,68,69,6a,6b,6c,6d,6e,6f,70,71,71,72,\
  73,74,75,76,77,78,79,7a,7b,7b,7c,7d,7e,7f,80,81,82,83,84,85,86,87,88,88,89,\
  8a,8b,8c,8d,8e,8f,90,91,92,93,94,95,95,96,97,98,99,9a,9b,9c,9d,9e,9f,a0,a1,\
  a2,a3,a4,a5,a6,a6,a7,a8,a9,aa,ab,ac,ad,ae,af,b0,b1,b2,b3,b4,b5,b6,b7,b8,b9,\
  ba,ba,bb,bc,bd,be,bf,c0,c1,c2,c3,c4,c5,c6,c7,c8,c9,ca,cb,cc,cd,ce,cf,d0,d1,\
  d2,d3,d4,d5,d6,d6,d7,d8,d9,da,db,dc,dd,de,df,e0,e1,e2,e3,e4,13,14,15,16,16,\
  17,18,19,19,1a,1b,1c,1c,1d,1e,1e,1f,20,21,21,22,23,24,24,25,26,26,27,28,28,\
  29,2a,2b,2b,2c,2d,2d,2e,2f,30,30,31,32,32,33,34,35,35,36,37,37,38,39,39,3a,\
  3b,3c,3c,3d,3e,3e,3f,40,40,41,42,43,43,44,45,45,46,47,47,48,49,4a,4a,4b,4c,\
  4c,4d,4e,4e,4f,50,50,51,52,53,53,54,55,55,56,57,57,58,59,59,5a,5b,5c,5c,5d,\
  5e,5e,5f,60,60,61,62,62,63,64,65,65,66,67,67,68,69,69,6a,6b,6b,6c,6d,6e,6e,\
  6f,70,70,71,72,72,73,74,74,75,76,76,77,78,79,79,7a,7b,7b,7c,7d,7d,7e,7f,7f,\
  80,81,81,82,83,84,84,85,86,86,87,88,88,89,8a,8a,8b,8c,8c,8d,8e,8e,8f,90,91,\
  91,92,93,93,94,95,95,96,97,97,98,99,99,9a,9b,9b,9c,9d,9e,9e,9f,a0,a0,a1,a2,\
  a2,a3,a4,a4,a5,a6,a6,a7,a8,a8,a9,aa,ab,ab,ac,ad,ad,ae,af,af,b0,b1,b1,b2,b3,\
  b3,b4,b5,b5,b6,b7,b7,b8,b9,b9,ba,bb,bc,bc,bd,be,be,bf,c0,c0,c1,c2,c2,c3,c4,\
  c4

Here's hoping that this gets you even closer to your solution, sanlyn.

-vhelp 5189

[sanlyn]

Put nearly a full day yesterday into looking for anything that appears to be LUT data in the WinXP registry, but found nothing -- at least, nothing that I could say was safe to use. As I suspected, Win2K and later uses color management differently than earlier Windows. Apparently apps today call DX transforms, Win98 is different.

It might be possible to make a copy of sRGB .icm, then copy the binary data from a "real" .icm into that and name it sRGB. But that's potentially dangerous, pretty much like fiddling in a tv set's service menu without knowing what you're doing (a definite no-no).

Still trying to find a few spare days to get Visual C++ 6 into my two video PC's for access to the multimedia SDK (earlier VisualStudio 5 has old VFW API's, which will go on my ancient Dell laptop). And the java SDK.

vhelp, your idea of an interface or plug-in similar to VLC Player's image control is another way to go, using java to modify VLC's open source. You could play a test-pattern video such as GetGray thru VLC and use a colorimeter and HCFR on another PC to set values for that control. For an idea of how this is done, see here:
http://www.curtpalme.com/forum/viewtopic.php?t=10457
This webiste is about tv's and projectors, but HCFR can read sRBG on a PC system as well as other color spaces. Or mybe java can connect to .icm's like other apps.

That means a long-term project that properly belongs in the programming forum.