[Openicc] Printing targets: App or driver ? Profiling RGB or CMYK

Chris Murphy lists at colorremedies.com
Tue May 17 18:39:18 PDT 2011

On May 17, 2011, at 6:58 PM, Robert Krawitz wrote:

> On Tue, 17 May 2011 18:15:22 -0600, Chris Murphy wrote:
>> When I suggest DeviceN not possible, I'm referring to the smaller
>> consumer inkjets. I don't know that there is actually a DeviceN
>> path. I know there is one for the 17" and wider Epson printers. But
>> for the R1900, R2000, WorkForce printers, Picture Mates, etc. I
>> don't know that they accept anything other than sRGB. Same for the
>> $50 HP's and Canon's.
> Well, *I* do, at least for Epson printers.  *All* of the current
> Epsons, bar none, have a DeviceN path.

OK good to know.

> It's possible that the printers have individual calibration tables; I
> don't know.  It might be interesting to put a USB tap on the
> bidirectional communication and analyze what flows in each direction.
> The sRGB path (ESC/P-R) is pretty straightforward.
> But I don't know about this "repeatability between prints" thing.
> Unless there's evidence that there's a problem with, say, Gutenprint
> but not with the OEM driver, it sounds like FUD.  One test would be to
> take a look at print to file with the OEM driver that's then sent to
> the printer vs. print directly to the printer.

I've done this test myself with the OEM driver versus various RIPs that either use Epson's HTM, or implement their own driver. Only the OEM consumer driver has this consistency such that the same profiles really do work quite well between printers. Without it, absolutely for proofing, probably for more discerning photographers, per printer custom profiles are necessary.

Chris Murphy

More information about the openicc mailing list