Re: Colorsync-users Digest, Vol 3, Issue 181
Re: Colorsync-users Digest, Vol 3, Issue 181
- Subject: Re: Colorsync-users Digest, Vol 3, Issue 181
- From: Mike Strickler <email@hidden>
- Date: Sat, 20 May 2006 15:26:10 -0700
Message: 6
Date: Sat, 20 May 2006 09:42:47 -0700
From: Dana Rasmussen <email@hidden>
Subject: Re: arbitrary profiling targets
To: Roger Breton <email@hidden>, Roberto Michelena
<email@hidden>, Marco Ugolini <email@hidden>
Cc: Colorsync <email@hidden>
Message-ID: <C0949517.2BFCA%email@hidden>
Content-Type: text/plain; charset="US-ASCII"
That's how some people build profiles from measurements of over
10,000
points, instead of your typical 1000-2000 measurement points.
Have those people ever reported obtaining better results through such
"extensive" profiling targets, as opposed to using the more popular
mundane
targets like the 928 patches IT8.7/3 or the 1485 patches ECI2002 or
some
proprietary ones (PrintOpen 210 basic target-type, PMP TC35 400 some
patches, etc)? For well-behaved CMYK devices? I understand RGB output
devices benefit from such extended sampling techniques for which the
name of
Atkinson immediately comes to mind. But, for what I call well-behaved
devices like well-maintained or new sheetfed presses, I would tend to
think
that such extensive sampling techniques are a bit of an overkill.
Personally, I have need any studies in the litterature to this effect.
Regards,
Roger Breton | Laval, Canada | email@hidden
http://pages.infinit.net/graxx
This is a question I have too. And even in respect to RGB devices.
I would think at some point the software creating the profile is
saturated
with data. While I can demonstrate that, at least for a wide gamut RGB
printer, I can build a better profile with a 918 patch, than a 283
patch,
and maybe 3600 is better than 918. Canon is using 69000 patches to
create
the profiles for the icf5000. Can the difference be seen?
I would be very curious to hear from someone who actually understands
the
math.
--
Dana Rasmussen
Seattle, WA
All color management calculations are founded on interpolation. The
idea is that the response of the device follows some more ore less
smooth curve. The less smooth the response is the more sampling points
are needed. So long as the difference between the real output and the
predicted, interpolated output for a given input signal is less than
can be detected visually it shouldn't help to add another patch. It
would be a very strange device that required 69000 samples to make a
good profile. And by the way, has anyone who is concerned about this
worried about the mere 30-50 patches used by their RIP for
linearization?
Mike Strickler
MSP Graphic Services
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden