Graeme,
I found a problem with the program I used to verify that the scanin
values made sense (they did not). I guessed that if there was something
wrong with scanin it would have been known by now, so I used 2-3 other
programs just to make sure the scanin data made sense and now I get a
very good match. Sorry about that.
Still, I'm using a ColorChecker Passport 2 instead of a ColorChecker and
I'd like to verify the reference are an excellent match or better yet
create a new set of reference files CCP2Half.cht, CCP2Half.cie. I'm
using an I1 Pro 2, should be good enough to get good data, right?
Currently, I'm not sure I'm getting a good match with ColorChecker.cie
and my readings of my Passport with my spectro. What kind of say DE2000
I should expect?
Also I see the ColorChecker.cie doesn't have spectral data, would there
be any benefit in having them?
I intend to measure my Passport at least 5 times and average the
results. Would that be enough?
My goal is to make sure I don't feed bad data to colprof or whatever
program I'll use to make my profile. I see in some papers they seem to
obtain much better results (avg. DE2000) then I got so far with colprof,
(matrix only) it's the main reason I started to verify I got good data
to use with colprof.
Any suggestions on making reference files for the full set of CCP2 patches?
With PS, you can display 0-255 (8 bit), 0-32768 (15 bit PS anomaly ?)
and 0-1 (32 bit) and same for CYMK.
On 2/28/2021 8:13 PM, Graeme Gill wrote:
Note that by convention Argyll stored device values as %, since this
is independent of any particular binary encoding. Many other popular programs
use
a mixture of encodings, such as 0 to 255 for 8 bit RGB values, and % for
CMYK printer values. I'm not sure whether Photoshop shows 0 to 255, 0 to 65535,
0 to 255.996 or 0 to 100 for 16 bit RGB values.
Cheers,
Graeme Gill.
