Differences in color between various lamphouses are caused mostly by the reflectors, heat shields, or other components in the optical path. Dichroic coatings and the like certainly do age and can affect color temperature when they do.

Regarding your specific question, were you able to obtain a 6500K reading at any IRE level? If you were, but getting 5100K at 30 and 80, you have some gray-scale tracking issues to deal with. In a machine meant for professional use, there should be enough range in the RGB gain, gamma, and black level controls to achieve a consistent color temperature reading at any level from near 0 to 100 IRE. If your color temp "curve" was fairly smooth, adjustment of these controls may yield improved results. If your curve had some notches in it, that would indicate aging or dirt problems somewhere in the optical path. How smooth was your curve?

To what degree does the incandescent glow of the hot electrodes contribute to the C.T. of the light?

Probably little, if any but I was just wondering wheter one had to let the lamp get "hot" before making any CT measurements.

Also, if you've focused a lamphouse with a double reflector, the electrodes in a poorly focused lh look like shadows in comparison to the light from the gas ball. Still, that anode gets awfully hot.

As Paul Mayer notes, gray scale "tracking" is important for any electronic display, to avoid having a color shift as you go from highlights to shadows. The absolute color temperature is not as critical as good gray scale tracking, since the eye will quickly adapt to minor variations in overall color temperature.

FWIW, "daylight" is approximately 5500 degrees Kelvin, and is a mix of sunlight and bluer "skylight". I suspect human vision has evolved optimized for daylight. Xenon and carbon arc lamps provide light that closely approximates daylight.

------------------
John P. Pytlak, Senior Technical Specialist
Worldwide Technical Services, Entertainment Imaging
Eastman Kodak Company
Research Labs, Building 69, Room 7419
Rochester, New York, 14650-1922 USA
Tel: 716-477-5325 Cell: 716-781-4036 Fax: 716-722-7243
E-Mail: john.pytlak@kodak.com
Web site: http://www.kodak.com/go/motion

As for the whites going pink at some drive levels, that sounds like the kind of problem many CRT-based projectors have. Data-grade CRT machine designs had to resort to playing with the gamma in the blue channel in order to get enough light output at the upper IRE levels. Which made it impossible to get true grays at all levels (try telling that to a client that's become convinced of your color blindness! ). I don't know the characteristic gamma of DLP or D-ILA imagers or their drive circuits, but I would think that they have fewer problems than CRTs in this respect. It'll be interesting to see the results you get after the new lamp is in place and you've had a chance to play with the various RGB settings.

And regarding John's point about 5500K daylight, one of these days I'm going to read up on how the SMPTE settled on having two different standards for the color of white in the picture business--i.e. film being at 5400K and video being at 6500K. It really bugs me everytime I walk out of a screening room and into an ajoining telecine suite! One of these days...