Originally Posted by mark haflich
In my opinion, the bit about the Spectracal being referenced to NIST is absolutely irrelevant. It might be if one were not going to use individually generated offsets either by the distributor of the probe or by the calibrator in the field. In that case, the accuracy of the calibrators reference spectroradiometer is the thing that counts.
Xrite charges the OEMer slightly more for a NIST traceable Display 3.
I would be more careful about this question!
The usual procedure is to profile your 3 filtered + 1 unfiltered luminance meter = colorimeter by comparing the measurements of the Display Primaries with the results of a spectrometer (which measures the radiance intensity at numerous different wavelength points to interpolate the spectral distribution of the light and integrate the XYZ values according to the observer functions).
This gives you nothing else but a simple 3x3 matrix! The devices are only aligned at one given luminance level! I never saw any calibration software which offers you to create and use 3DLUTs for meter profiling.
So, your colorimeter must be fairly linear, or everything is broken! The gamma tracking of the calibrated display will be false. Moreover, your correction matrix will become invalid if the individual filtered channels have different non-linear characteristics.
If the colorimeter isn't linear enough then you need numerous matrices for numerous light levels. That would be a 3DLUT (like the Lab and XYZ cLUT based ICM profiles for printers and PC displays).
If the NIST certification grants you a trustfully linear sensor then it can actually be a MUST HAVE for those who want to trust in their profiled colorimeters almost as much as they trust in their reference grade spectrometers!
Sidenote_1: Of course, you can test the linearity of your colorimeter with your own reference grade spectro but that could be quiet a job and may worth a few hundred dollars for a busy man to get it done by NIST...
We shouldn't even talk about 3x3 matrix colorimeter profiles anymore because X-Rite showed us the way: they measure and save the spectral response of the colorimeters in the firmware! If you compare those with the spectral characteristics of a display (it wouldn't require you to measure every single display units all the time, you could just load the files of a given display model from your own or even from an online database -> you wouldn't need to carry your own spectro at all if you delt with the same display model or you have access to an online databse !!!) and BAAAM, your colorimeter has almost reference grade accuracy with it's own internal precision. (Linearity still remains a question though...)
SpectraCal and ChromaPure guys should update that branch averaged response data in the firmware for the PRO versions, put up an online database of spectral measurements from numerous display models and that's it.
The X-Rite product was indeed supposed to be revolutionary (just like the topic title says) but the miracle has been canceled guys, please return to your "good old" routines and think you make the real deal, thanks...
(Don't misunderstand me, it's still a nice sensor for it's price. I mean BOTH of them, even the C6! Only the software side is less revolutionary or even up-to-date than they want to think...)
Hmm... I am not sure if ChromaPure actually offers some kind of 3DLUT based meter profiles for their PRO labelled products. (They talk about "figures", not only matrices...)
If that's the case then a ChromaPure branded PRO version could be usufull for even those who have a spctro (unless... I don't know... does ChromaPure offer this kind of profiling on the field too...? I mean not only 3x3 matrix...?).