Quote:
Originally Posted by
PlasmaPZ80U 
Even though these 2 new standards won't be of any immediate use to me atm, it's nice know how easily new standards can be added to CalMAN 5.
Since you mentioned color matching function (cmf), can you explain what the selections for that in CalMAN 5 do? I believe it is for use with spectros?
Yes it is for spectros only.
The Color Matching Function or CMF is what turns the spectral data in to XYZ data. XYZ data is suppose to mirror the sensitivity of the three different cone types in the eye. The original 1931 CMF has some deficiencies in the way it accounts for near ultra-violet light. This is mainly because the technology they used to run their experiments didn't have a lot of near ultra-violet. The 10-degrees CMF like the 1964 CIE, are for large field of view color analysis and therefore aren't particularly relevant to video content. The newest 2-degree CMFs like the CIE 2007 CMF seeks to correct the issues with the 1931 CMF. But it didn't stop at the CIE 2007, and we have modified version of that CMF as well.
The problem that gets solved is the one where two displays read the same x,y,Y coordinates yet look different. A CRT and wide gamut CCFL almost always look different, even when they measure exactly the same on the CS-2000. Also Xenon bulb projectors rarely match standard UHP projection. When you switch to newer CMFs, you'll now get different x,y,Y values for these displays that were previously reporting a match. So you can do a much better job matching different display technologies with these newer CMFS.
The real catch to using any of these, is that since D65 has a spectral definition (it is not 0.3127,0.329, but a specific spectral output) so when you change the CMF, you change the location of D65. But the primary locations for rec.709, or any of the other standards don't have spectral definitions. The primaries are defined in the 1931 space assuming the 1931 CMF. So until we have new standards that account for using a different CMF we are stuck using the 1931CMF for anything other than trying to match to a LCD to a rec.709 calibrated CRT.
Of course a Colorimeter profiled off a spectro using a new CMF, would then take on the properties of the new CMF.
Edited by sotti - 8/29/12 at 10:12am