Post
by spacediver » 02 Jul 2014, 12:40
Unlike spectroradiometers, which measure radiant energy at specified intervals across the visible spectrum, colorimeters use filters that approximate the standard observer functions. Because it is very difficult to match these functions perfectly, colorimeters are generally not as accurate (although an ideal colorimeter with perfect filters would be more accurate than any spectroradiomater). The tristimulus error of a colorimeter depends on the interaction between the way in which the filters differ from the standard observer function, and the spectral signature of the display primaries, and colorimeters therefore use things like offset tables to compensate for these errors. A colorimeter that has good offset tables for a particular display will be extremely accurate on that display.
The DTP-94 has factory offset tables that are designed for the phosphors used in BVM CRTs which are pretty much the same phosphors as those used in high end trinitrons, and very close to most other modern CRTs.
When you use a DTP-94 on a display that has a different set of spectral signatures (even if it has an identical gamut to the BVM CRTs), you have no guarantee about its accuracy. Remember, the same chromaticity can be derived from an infinite number of spectra.
I'm not sure how the spectral signatures of the CCFL display used in the luminous landscape tests compare to the vg248qe, but if it's close, then a DTP-94 will do a decent job (< 5 dE), although it will be more accurate with a CRT.
If anyone in Toronto has a vg248qe, I can come measure the primaries with my DTP-94 and my i1 pro (spectro) and see how accurate the DTP-94 is.