Both Windows and Firefox are fine with XYZ LUT profiles created by Profile Maker, ColorEyes, basICColor, ect. Argyll is the only one that doesn't work.
Are you sure about that ? (see below)
As stated previously I believe it may be because Argyll doesn't include the rXYZ, bXYZ, gXYZ, rTRC, bTRC, and gTRC tags on it's XYZ LUT profiles like the other previously mentioned applications do in their XYZ LUT profiles. It seems like it should be a simple fix for Graeme Gill to make if that is really all it is.
The tags you quote are used just for matrix profiles. If you remove those tags from the profiles so that only the XYZ cLUT tags remain, do they still work ? (I bet they won't). ie. what you are seeing is that some applications fill in both cLUT and matrix tags, and then the CMM will choose what to use. It's hard to know which profile type is actually being used though, and in this case it seems that you think the XYZ cLUT is being used, but in fact the XYZ matrix is being used. Using Argyll profiles it's clear what is being used and what is not.
Yes, both Windows and Firefox have always been fine with shaper & matrix profiles created by Argyll, but I want to be using an XYZ LUT profile, which is why I posted this.
It sounds like this is not possible because Windows XP doesn't handle them.
Why is using matrix profile for color critical work more appropriate then a higher quality LUT profile? Unfortunately the Argyll shaper/matrix profiles don't produce close similar enough enough output to the LAB LUT profiles created from the same .ti3 values, which I why I just want to use the same XYZ LUT profile in Photoshop, Windows, and Firefox so they all behave near identically (or at least close enough).
Matrix profiles can give much smoother results, but will only fit well to devices that behave quite additively. cLUT based profiles are far more flexible in fitting to other device behavior, but being table based have a degree of piecewise linear table type artefacts, and are therefore not quite as smooth. Graeme Gill.