Re: Profiler RGB
Re: Profiler RGB
- Subject: Re: Profiler RGB
- From: email@hidden
- Date: Thu, 10 May 2001 10:47:16 EDT
In a message dated 5/10/01 5:40:13 AM, email@hidden writes:
>
A question about ColorVison's "Profiler RGB " if I may. I can't
>
just understand how it works from the scanner - it comes as a
>
printer testfile - on disk - with no printed target and is said
>
to be able to use an unprofiled scanner to read the patches once
>
printed.
>
>
I am sure it works fine because I have a lot of respect for
>
ColorVision but to me it seems like this:
>
>
The printed result likely contains distortion of the image
>
appearance, in relation to the original file data, so the
>
Profiler RGB SW plainly needs to make an output profile which
>
provides for correction of this deviation. That bit's obvious,
>
and easy.
>
>
I just can't get my head around how the printed target gets
>
scanned.
>
>
If the unprofiled scanner distorts the printed image's appearance
>
[ie it adds to or subtracts from the printer errors ] - then
>
obviously that distortion should not become part of the printer
>
[output] profile correction. So how can the Profiler RGB SW
>
"know" whether it's looking at scanner distortion or printer
>
distortion, or a combination?
>
>
I understood how Praxisoft [WYSYWYG?] did it with their pre -
>
printed target which profiled the scanner first - but this seems
>
very difficult to understand.
ColorVision is unlikely to tell you how its done, but comparitive testing
proves it works about as well (and more quickly and simply) than the
tape-on-an-IT8 method.
After all, the IT8 has a white patch that's not white, a black patch that's
not black, a gray ramp that's not neutral, and a data file thats not very
accurate to the actual target, due to mass production. The extrapolations and
approximations required to account for all this is not trivial either. And in
the end you have defined your scanner's response to a photographic target,
while what you are trying to read is typically inkjet output. So you have
metamerism, gamut, and white point issues added at this juncture as well.
So maybe that secret sauce used to read the target without an IT8 is not much
more of a stretch than all the factors the IT8 introduces...
C. David Tobie
Design Cooperative
email@hidden