Spectral Response and Colorimeters....
Spectral Response and Colorimeters....
- Subject: Spectral Response and Colorimeters....
- From: "tlianza" <email@hidden>
- Date: Sat, 2 Feb 2002 08:59:55 -0500
As someone who has designed color instrumentation for 25 years, I thought
that I would weigh in on this one...
There are a number of methods that can be used and used quite well...A brief
outline,
1. No Filter
If the phosphors are well known and controlled and the spectral response of
the instrument is well controlled, you can build a monitor calibrator that
will work quite well with that monitor with a very simple calibration. The
calibration must be able to determine correction factors to measure the R,G
and B luminances correctly. The upside to this technique is that is cheap,
it works very well for a single display type, and there isn't much that can
go wrong in the field. The downside to this method, it requires separate
measurements of the Red Green and Blue components and the montior
electronics may interact and the combination of the RGB guns may not be
linear due to power supply sag or other factors. This was the basis of the
original Super Mac SuperMatch calibrator. That calibrator was a great
example of how simple technology could do the job quite well. I believe
that it was done by a team at Tektronix
2. Arbitrary Red, Green, and Blue or any other colors for that mater.
If the phosphors are well known and the spectral response of the instrument
is well controlled, this mechanism will allow you to make very accurate
measurments of the color on the screen without turning on the individual RGB
guns. With proper calibration this mechanism works as well as any
spectrally qualified device. I once had a technician at standards lab
accuse me of "hardwiring" the calibration to his $40,000 specro radiometer.
What he didn't realize was that my in-house standard was calibrated by his
device, and the "cheap" calibrator was actually very linear and calibrated
to that standard. If we moved this inexpensive device onto a completely
different phosphor set, the results could be different.
3. Correlated Color Matching function response.
I've published a number of papers that mention this technique because it is
used in some of our products. If you know the nature of the sources that
you are going to measure (i.e. a number of different phosphor sets) you can
design color matching response filters to closely match these families of
phosphor sets in the regions of greatest correlation. This maximizes the
accuracy of the measurement for these particular families, at the risk of
making larger errors for non standard sources. This method can be shown to
ALWAYS beat the accuracy of a spectral device of bandwidths greater than 15
nm (total). This is particularly true for devices that have varyng bandwidth
with wavelength (i.e. all array based instrumentation). The correction for
such spectral devices requires and operation called deconvolution and it is
quite difficult to perform.
4 Orthogonal Color matching response.
The spectral response of the instrument need not preciesly match the color
matching functions. These functions can be derived by a linear weighted sum
of the instrument response functions. Instruments based on this philosophy
will measure a wide range sources with fairly good accracy (often better
than a spectral device) and then potentiall fail fantastically on a
particular narrow band source such as sodium vapor or mecury lamps
The point to all of this is that there is no dominant technolgy that works
well in all cases. People who measure emissive sources for a living tend to
use colorimetric devices made by companies such as LMT in Germany. Spectral
devices when used by people with good training and understanding are
perfectly applicable to most tasks as well. We all design instruments to
meet the cost, speed and accuracy requirements dictated by the application.
Tom Lianza
Technical Director
Sequel Imaging Inc.
25 Nashua Rd.
Londonderry, NH 03053
email@hidden