I've been repeating this with other testcharts and still no luck, softproofing is still wrong.... Is there some way I could fix this? Or some other way of applying the calibration?
Kind regards, Ridouan. On 15-01-2012 05:07, ridouan wrote:
Hi,I've been experimenting with calibrating a 'regular' inkjet (Canon MP) using printcal cs. After some initial headaches I've been able to get almost the whole process running correctly, with one slight bump at the end: using applycal to alter the curves of the ICC profile.To clarify a bit, these are the steps I've taken:1. Generate a suitable target for calibration (targen -d2 -s64 -e3 -f0) and generating a tif. 2. Print this target without color management and no calibration (as per the online docs). 3. Read the patches, create a new .cal file from the ti3 file (printcal -v -p -i) 4. Generate a new target for profiling (targen -d2 -G -g0 -f416) duplicating this into two sets of data (chart-std and chart-cal). 5. For both sets I created the .ti2 and .tif files using printtarg (printtarg -iSS -R0 -t chart-std and printtarg -iSS -R0 -t -K printer.cal chart-cal), visual inspection shows a color difference between the two tifs. 6. Printed both TIF files and read them, creating two profiles "chart-std.icm" and "chart-cal.icm" (colprof -qm -cmt -dpp -SsRGB.icc chart-std and colprof -qm -cmt -dpp -SsRGB.icc chart-cal). 7. Applied the calibration curves to the "chart-cal.icm" profile, resulting in the final profile, "chart-cal-applied.icm" (applycal -a printer.cal chart-cal.icm chart-cal-applied.icm). 8. Finally, I installed all 3 profiles to the system32\spool\drivers\color directory.I've examined the profiles (gamut volumes, profile curves, avg. dE, etc.) but since that doesn't say much about the final output (it's all about visual output after all :)), I used Photoshop to check and print a test image. That's where the slight bump is. Printing with the standard chart-std.icm profile (no calibration involved) everything is as expected: softproofing is OK and the print is also OK. With the second profile (chart-cal.icm: calibration involved, NOT applied), the softproofing is OK, but the final print shows way too much yellow. With the third profile (chart-cal-applied: calibration involved, also applied), the softproofing is wrong (too much blue), but the final print is OK (visually almost identical to the output of the first profile).Just to verify, I did a invprofcheck of all three profiles, with the following result:chart-std.icm: errors: max. = 7.738869, avg. = 1.191317, RMS = 1.528485 chart-cal.icm: errors: max. = 5.860738, avg. = 1.403858, RMS = 1.668143chart-cal-applied.icm: errors: max. = 38.386883, avg. = 13.736625, RMS = 17.136439To me IMHO the B2A / A2B tables are not processed equally... Am I missing some step here (BTW, I tried to use revfix, but it accepts CMYK profiles only...)? HELP! :)Kind regards, Ridouan.
