Re: CIE document for tristimulus weight calculation


Gernot Hoffmann wrote:
Jim,

my interpretation of the OP's question is this:
He has already the CMFs with sufficient accuracy.

The interpretation of the spectra SAMPLES depends on
the method how to reconstruct a continuous function
which can be integrated analytically.
E.g. a set of sample values can be integrated by Euler=
Rectangle, Trapezoid=Linear, Simpson=Quadratic,
Natural Spline or whatelse.
The weighting functions are assigned to different
interpolation methods (guess about the continuous
function).

This was IMO explained by Danny Rich, but on the
other hand there seems to be some information missing:
the bandpass characteristic of the spectrophotometer.
Without knowing this, the choice of an optimal inter-
polation is uncertain.

Best regards --Gernot Hoffmann

Yes of course. Publication 15:2004 recommends that the bandpass be the
same size as the sampling interval. Thus for 1nm intervals the
bandpass should be 1nm and for 5nm intervals the bandpass should be
5nm. There is some disagreement about this restriction from lighting
engineers - mostly with respect to readings of incandescent sources or
the broadband emission of older phosphors. The CIE recommendation on
CRT colorimetry clearly demonstrated that narrow-band phosphors will
not be characterized accurately with a wideband spectroradiometer no
matter how closely spaced the sampling interval.

Danny

.

Relevant Pages

  • Re: Swift Data rules out beamed theory
    ... the spectra was before it was reinterpreted to support ... I said your reason for dismissing the ... redshift interpretations of the data were made under the ... How would an alternative interpretation provide ...
    (sci.astro)
  • Re: what is matter?
    ... red color actually red? ... is our mind's interpretation of a particular slice of the ... any objective sense, ... thing as "spectra" in any objective sense. ...
    (talk.origins)
  • Power Spectrum
    ... If I have two spectra of identical shape and hence the same area but ... one is say 6dB lower than the other, what is the interpretation of ... The spectra are both band-limited via a bandpass filter. ...
    (comp.dsp)