[argyllcms] Re: Printer calibration based on reference print
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Tue, 14 Mar 2017 12:33:17 +1100
C. H. wrote:
Hi,
my goal with printer calibration in Linux is to get similar photo printing
quality as my
printer provides used with Windows. The problem is that the printer does
toally fine in
Windows, but in Linux with the available Gutenprint driver I have issues with
the color.
It's going to depend a lot on what mode the Gutenprint driver is using with
the printer. If it is using "RGB" mode (i.e. using the printers separation
tables), then the profile is going to be relatively well behaved. If it
is driving the printer in a more raw mode (i.e. CMYK etc.), then the is a lot
more to setting a printer up, such as establishing per channel ink limits,
per channel linearisation, total ink limits, optimal black separation etc.
The latter takes a good deal of experience to do quickly and well.
My idea here ist that I do not need to have a calibrated scanner (avoids to
buy a
professional target) as I just want to compare the print results from my
printer in linux
and windows. Please correct me if I am wrong.
I doubt that will work very well. Putting aside the practicalities of it
(trying to create a profile that when linked with a source profile
produces the same scanned value) :- if the printer uses a different
ink combination to produce a color, the resulting spectral reflectance between
the two systems will be different, and the scanner won't "see" the color the
same way you do, so it's "corrections" will be wrong.
3. I scanned the resulting prints with my uncalibrated scanner
However, the resulting profile did not solve my color problem (mostly problem
with the
grey ramp).
That's to be expected. A Scanner is not a colorimeter. The only way it
can come close is to calibrate it using the same media as you will
be profiling, leading to a Catch-22, since you need a reference
for your scan chart made using a real instrument.
Does my approach make sense?
No. You aren't going to get far without a reliable reflection reading
spectrometer. A scanner won't substitute, even if you were using
an IT8 target to profile it.
Do you have any recommendations or suggestions?
Either buy or borrow a spectrometer. The cheapest (currently available new)
one out there that can do the job is the X-Rite Colormunki Design or Photo.
Even a second hand (but working!) i1Pro rev A would be better than
using a scanner.
Cheers,
Graeme Gill.
Other related posts:
