Apple

On Aug 13, 2005, at 8:00 AM, Roger Breton wrote:

It is very common practice to use a tungsten light source for the
illuminant in spectrometers. The light source spectrum is removed from
the measurement to give the spectrum for the material only, which can
then be combined mathematically with the D5000 theoretical spectrum to
give a result for the object as if it were illuminated with D5000
light. There are no illuminants that produce the exact D5000 spectrum,
so this is common practice.

Robin Myers

In theory (and practice), you say that the tungsten light source is used
*only* to derive the spectral radiance factor, and then that information is
integrated with the user's choice of CIE or Custom measured Illuminant and
Observer function to compute the tristumulus values?

Yes. This is the way all the visible spectrometers I have used operate.

Makes sense to me. What's interesting, nevertheless, is that we'll obtain
different radiance factor depending on the actual light source used in the
instrument, won't we? Or is this difference attenuated or eliminated when
taking into consideration traceability and factory calibration?

The instruments are factory calibrated to standards that account for the light source. In the US the standards are NIST standards. This method of operation works for most colorants, the exceptions are ones that fluoresce. Then, the amount of fluorescent excitation energy in the illuminant may be different for the spectrometer's illumination and the desired result illuminant. However, these are not encountered often, with the exception of fluorescent whitening agents, which is a large subject by itself.

Regards,

Robin Myers

email@hidden

References:
	>Re: DTP70 specs (From: Roger Breton <email@hidden>)