[argyllcms] Re: Scanner utility that significantly improves colorimetric accuracy - printer profiles
- From: Martin Gray <mdgray@xxxxxxx>
- To: "argyllcms@xxxxxxxxxxxxx" <argyllcms@xxxxxxxxxxxxx>
- Date: Tue, 11 Aug 2020 19:09:08 +0000
Does this mean that we could use a flatbed scanner to create accurate paper
profiles instead of a spectrophotometer, I mean after the initial calibration
scan?
Emanuele
Yes, and no. I tried this last year printing a single US Letter page of 2,000,
4mm patches. The printer profile that resulted was quite good at .6 dE ave.
error and only a few tenths of a dE worse than a custom profile on the same
printer. And it made pretty decent profiles for other printers. Using the same
scanner profile with different printers to make printer profiles, ave dE went
up to about .9 with glossy and 1.1 with matte. OBA presence had minimal impact
as expected since everything was done with M2. Scanners have no significant
uV.
But while it eliminates almost all the reflection induced errors, it does
nothing for metameric failure. The Epson V850 exhibits large metameric errors
between scans of printed material (CMY Based) and chem. developed photo prints
with the IT8 chart Epson provided is
For example the yellow-orange loc(6,2) of an XRite Colorchecker and the closest
similar color in the IT8 or lab(chem) photos have similar spectra but shapes
all are quite different from the spectra of the same color when printed with
CMY printers. However, all the printers and other printed material with similar
colors exhibited CYM spectra patterns. On the V850 the metameric failure on
this color is on the order of 8-9 dE. How much metameric error occurs is a
function of the scanner’s illuminant and its color filters.
So, for the most accurate scanning of printed materials, in addition to using
the reflection correction utility, a scanner should be profiled using a
printed patch set. For accurate scanning of chem. photos, the scanner should be
profiling using an IT8 on chem paper.
doug
Other related posts: