[argyllcms] [argyllcms] Re: bin/average: averaging and possible outlier eliminationfor three or more .ti3 sets?

  • From: Nikolay Pokhilchenko <nikolay_po@xxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 31 Aug 2009 18:42:50 +0400

Graeme Gill wrote:
>
> When you say "misread", do you mean misread without noticing the problem, or
> do you mean there is a reported misread ?
>
> I haven't seen any examples of non-reported misreads, with the exception of
> the direction detection problem I mentioned before. I'd be interested in
> hearing about any other reproducible examples of non-reported misread 
> problems.
>
> Graeme Gill.
>

I have read by chartread about 50 difficult charts from non-linearized inkjet 
printers and thermo-sublimation prints. I've got error messages from almost all 
of it. But only 3-4 times (6-7 strips) I've got no errors in bidirectional mode 
while direction determination error have took a place. The cases was confirmed 
by manual patch-by-patch readings. The patch number in row was 12-18 in that 
error cases. Some of charts was without spacers.

It must be mentioned that preliminary profiles for targen in my error cases was 
rather far from real print (average dE~11, max dE>35). And errors haven't took 
a place while preliminary profile was quite good (max dE<10) and the chart have 
a spacers.
Another minor issue for printtarg-targen is bad bath-to-patch contrast in 
dark-green and dark-brown tones without spacers. If I have a good preliminary 
profile with max dE<5 and do the chart without spacers, the chartread (or 
eyeOne instrument) cant distinguish a transient between adjacent dark patches. 
I had to scratch the paper between the patches to get the strip reading. The 
printtarg didn't know exactly what patches transient can be read robustly by 
EyeOne.

Other related posts:

  • » [argyllcms] [argyllcms] Re: bin/average: averaging and possible outlier eliminationfor three or more .ti3 sets? - Nikolay Pokhilchenko