[argyllcms] Re: Printer calibration - curves
- From: ridouan <ridouan@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Wed, 18 Jan 2012 10:46:20 +0100
On 18-01-2012 03:11, Graeme Gill wrote:
To be fair, it's not clear whether print calibration curves aid an
"RGB" printer or not. The facility is really aimed at the situation
where the device space is native (ie. CMYK) or close to native (CMYK
are composed of light + dark inks into a composite curves of CMYK).
Well, initially I wanted to use the calibration feature to see whether
it could help me skip selecting a 'Media Type' in the printer settings
and get correct ink output irrespective of media type selection. After
seeing the option 'post-linearization' in PM5, I started wondering
whether the calibration could also be used as an 'after-deployment
finetuning' of profiles.
I'm not sure whether this is a utter bullocks or not, but lets say that
I have 10 papers I've created profiles for (for each paper 3 or 4 A4's
full of patches) in the dead of winter (minus 10 C, lots of rain) and
windows open. Would the profile be just as valid six months later when
the temperature would've reached 35 degrees C? Would it make sense to
print a linearization chart on a specific paper (let's call it paper A)
6 months after initial profiling of said paper, create a NULL
calibration with the initial profile, use printcal and the linearization
readings to create a delta that with the help of printcal results in an
updated profile for paper A that's refreshed, as if I just created it
with a full set of patches? For one paper full reprofiling would not be
such a bother, but if you have a lot of different papers/profiles, I
could imagine it starting to get tedious after a while...
Second option I was wondering about is whether a delta for a specific
paper would be usable to refresh profiles for other papers (but same
printer/ink combination). In theory I guess it should be possible (since
the only difference would be the change in ink output, all other factors
such as paper and printer remaining the same) and thus the resulting
calibration curve would purely describe the drift in ink output, which
should then be just as applicable to other papers.
Last option would be to use it as an ongoing quality control process: it
would be fairly trivial to regularly print a lin. chart and examine the
resulting curves to determine (possible) drift in output, I'd imagine
that if you were going start a big panorama print that'd eat half of
your ink cartridges you'd like to know beforehand if there's any reason
to maybe recalibrate, instead of afterwards.
The rationale for choosing the linearization option instead of a small
amount of patches would be that using the linearization chart results in
an independent examination of each ink channel, without any
'crosscontamination'...
Would this make any sense (for an RGB printer)?
And this confirms it.
It should be easy enough to add the calibration to the A2B tables too, to
fix this problem. What platform are you running on ?
Graeme Gill.
Currently running on Windows 7 x64...
Kind regards,
Ridouan.
Other related posts:
