[argyllcms] Re: eye-one pro Rev A/Rev D IR calibration?

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 30 Jun 2010 22:44:31 +1000

Klaus Karcher wrote:
I noticed some weird differences between an eye-one Pro rev. A and my rather new rev. D: the reflectance measurements in last few bands (approx 743 ... 750nm) are significantly lower with the rev. D than with the rev. A. Furthermore the results for this bands seem to be more temperature-sensitive.

I guess that's hardly a surprise, since the i1pro doesn't officially
go below 730nm. Anything below that is an unofficial extension. Your
experience seems to indicate that there are valid reasons for not
going beyond 730nm. (Perhaps the Rev D uses a slightly different
diffraction grating that is less efficient at longer wavelengths than
the one fitted to the Rev A.)

My reaction would be to curtail the extension to 740nm.

reflectances rev D divided by reflectances rev A, mean of 3 bands,
(443...450 and 743...750nm) ploted against patch index.

I guess the issue has practically no influence on the Lab values as the spectral sensitivity of the standard observer is close to zero in this region, but I need reliable data also in this region. Any ideas how to improve the IR results of my Rev D?

Do you really need to go below 740 nm ?

The thermal drift is a problem without extra capabilities in
the instrument (such as a sensor temperature, to be able to
calibrate the black level). I guess noise could be improved by averaging
more readings together.

(You might like to try the development snapshot, to see whether the
changes to the strip reading code affect your numbers at all too.)

Graeme Gill.

Other related posts: