[argyllcms] Re: Profile for screen printing problem

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Fri, 16 Jan 2015 17:01:06 +1100

Semund Svelle wrote:

> The resulting profiles have some strongly inverted points. I have repeated 
> the test with
> smaller patch amount but same problem. Could it be that the screen frequency 
> is to low
> (60lpi) so that the i1 can't read properly (printtarg -a1.5)?Is there please 
> anyone with
> experience from profiling screen printing that could help me?

The calibration curves appear to be pretty reasonable, and don't appear to
have any particularly poor behaviour.

The profile fit shows good average errors, but a high maximum error
(max. = 19.092950, avg. = 1.048334, RMS = 1.509504)
It's possible this is due to some patch mis-reads or inconsistency
in the patch reading.

Removing some of the largest delta E fit patches (E14, E15, E19, V21, T18, AN20)
seems to remove some of the lumpiness in the profile. Decreasing the -q
and/or increasing the smoothness factor (colprof -r 1) could also help.

Some ways of dealing with a coarse screen:

Use more test points (i.e. 2000 or 3000).

Expanding the patch size (targen -a scale) and also
then reading the chart several times and combining the readings
(average -m) may help.

Graeme Gill.

Other related posts: