Re: Calibration targets for scanner / Calibration of CUPS color laser printer (follow-up)

[Richard Hughes <hughsient gmail com>]

2010/9/26 Jean-Michel Pouré - GOOZE <jmpoure gooze eu>:
> Do you think it is reasonable to believe
> that I can calibrate using this chain:
> * screen (using calibration tool)
> * scanner (using proof prints)
> * AND THEN printer (printing charts and then scanning them).

I believe you can (people have tried) but the resulting profiles are
so full of noise it's not worth the bother. You're doing complex color
maths on a reading (with an error) using a machine to read them (with
an error) and the resulting profile is going to be pretty crude. This
is why GCM doesn't support this method of operation.

That said, it might be profiles produced this way are better than
nothing, although lots of color Nazis would say that a 90% correct
profile is worse than no profile at all.

Richard.