[argyllcms] Re: Argyll i1pro 3nm spektral resolution

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 05 Nov 2007 11:24:29 +1100

Roger Breton wrote:
I can't help but marvel at the technical prowess of your device driver for
the i1pro. How on earth are you able to extract down to 3nm data out of this
device?

Well since X-Rite/Gretag don't provide Linux drivers for the i1pro, and
they're existing drivers don't have a compatible license for distribution
with a GNU licensed package, I've had to figure out how to understand
how the instrument works at the USB level. Luckily it has a reasonably
well organized interface, but what you get is the 1 dimensional CCD
sensor output from the instrument, plus the calibration information
stored in its EEPROM. (this is typical of how many low cost instruments
communicate.) This calibration information basically describes how to
resample the CCD data from the physical sampling of approximately
3.5nm (approximate because it's not calibrated) spacing to the
calibrated 10mn spacing. By interpolating this information,
and reformulating it using a sync filter template, I can create
a resampling into a 3.333nm spacing report.
The resolving power isn't really 3.333nm, but it does seem to
make a difference for spiky emissive spectra. Since I don't
have access to a suitable reference instrument, I'm not sure
though, whether it really improves it's accuracy or not.

Graeme Gill.

Other related posts: