[argyllcms] Re: RGB Printer profiling and ColorSavvy CM2C

  • From: Gerhard Fuernkranz <nospam456@xxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sat, 30 Jul 2005 19:19:38 +0200

Greg Sullivan schrieb:

I wrote:

My main concern at the moment is that saturated blues and magentas look too light and unsaturated.

The root cause might be that the device gamut is actually limited in these sectors of the color space. This is not necessarily a limitation of the printer, but can also be caused by the driver - you're using a RGB printer driver, don't you? Have you tried different driver settings? Are you using a "raw" print mode (like the NCA mode of Epson drivers)?

I also had weak blues and magentas on my inkjet prints in conjunction with the manufacturer supplied Linux driver (RGB only) and 3rd party inks and paper. First I tried to move up the cyan and magenta sliders in the driver settings. This helped a bit to increase the saturation, but as a negative side effect, the more I moved the sliders up, the device response became more and more non-linear and "rippled", such that the profile could not follow the device characteristics any more accurately. My breakthrough was eventuelly, when I changed the "print type" from "photo" to "graphics" in the driver settings. I don't need to boost cyan and magenta any more, nevertheless the gamut is now definitively larger than it was in "photo" mode, and profiling also works better with this print mode :-)

Unfortunately, for RGB printers, profiling cannot generally compensate all flaws of the color mangling done in the driver. First you need to find a print mode and driver settings, which deliver maximum gamut, but still result in a smooth device response, with gradients of less than 1dE per RGB unit at any location within the gamut. This can be a tedious try-and-error task, fiddling with various driver settings, which needs to be done manually first. Once you have found the optimal driver settings, you can start profiling ... But if you don't happen to find reasonable driver settings, then you are lost and profiling also can only partially compensate the deficiencies :-(

Regards,
Gerhard




Other related posts: