Re: UCR/GCR revisited
Re: UCR/GCR revisited
- Subject: Re: UCR/GCR revisited
- From: Roger Breton <email@hidden>
- Date: Tue, 10 Jun 2008 23:54:10 -0400
Ray,
> I suggest that you all purchase and download the interview with Bill
> Atkinson about his procedure for printing his beautiful book "Inside the
> Stone". You can find this interview on the Luminous Landscape Video
> Journal Volume 16. You can download this video at the
> "www.luminous-landscape.com/" web site.
I just did that. But after watching it, I thought it left me with more
questions than provided answers for. Sadly, I got the impression that
Reichman kept refering to "some other segment" where the things that really
interested me about Bill's color management technique would be covered.
Nothing in Vol17 gave me much new information over what I already knew about
the making of Within the Stone :(
> I suggest that a printer in North America step forward if
> he wants to repeat this "experiment".
I think a lot of NA printers would be well up for the challenge today. But
I'd imagine a lot of them wouldn't feel they have anything to prove either.
Too bad no NA printer was open-minded enough to help Mr. Atkinson at the
time. That's a shame.
> Here is a quote from the interview..."The pressman was surprised to find
> that all seventy two images in the book printed perfectly without
> adjusting the press for the content of each page." He just printed to
> the exact numbers that Bill had chosen from his press tests and
> characterization.
Yes, that's amazing. But that's nothing new, in a sense, compared to what we
know CM is supposed to do for us. Why does CM fails on us is a constant
challenge. It could be inks, paper but mostly press repeatibility. Or,
perhaps, why not recondiser the color model we're using? Or the optical
conditions under which our measurements are taken? I did not see anything
new or out of this world about Mr. Atkinson's account of his experience in
Vol17 DVD.
> Check out this video. It is a real eye opener to what can be done.
Again, I'm looking forward to that "other segment".
> In the interest of full discloser...Michael does an interview with me on
> Luminous Landscape Video Journal Volume 17. It is an interview about
> basic color science and not nearly as important as the interview with
> Bill that appears on Volume 16. I did not receive any monies for this
> interview and will not receive any financial benefit from the sale of
> any of these video interviews.
>
> If any of you do download and listen to this interview, I would very
> much appreciate your comments on what was said. Please give your
> comments on this forum.
Again, I did. For starters, I was giving you a few years less ;-)
One thing that annoyed me was how the interviewer kept adding his own bits
of knowledge over and above what you kept explaining all along. He was
asking you questions, bringing up issues but, in the end, he wouldn't
concede anything to his deeply held beliefs.
I like the part where you discussed the suitability of certain color spaces,
e.g. ProPhoto. Your argument about why go to something this big when the job
only calls for, say, AppleRGB, in terms of the size of color gamut required,
to hold the color of an image to its fullest extent. I could not agree more.
You argue that ProPhoto calls for end to end 16 bit workflow, otherwise it's
too risky. Again, I can't agree more. I like your explanation of what
happens to quantization, when dealing with "moderately size scene colors".
You argue that it'd be best to stick with color spaces that are more adapted
(smaller) to every day color imagery than ProPhoto. And I think it makes
perfect sense. Except, I once asked that very question to Kevin Spaulding at
Kodak, one of the inventors of ProPhoto, and his response was, nahh, 8 bit
does not cause any quantization with ProPhoto. True or false? Who's to
believe? I have my opinion, but.
Next, the interviewer goes about resisting your counter arguments to
ProPhoto. Citing things like butterflies, super saturated flowers. I'd wish
I had some ways to figure out what the typical quantization level would be
off some standard saturated and less saturated RAW images, to really debunk
this myth, once and for all. Problem is, like I said many times on this
list, there are no standard set of RAW test images that one could actually
use to test various workflows scenarios. So all these questions still remain
in the air for me...
Then you go on advocating that Adobe98 is, in your opinion, perfect for an
Epson printer as opposed to ProPhoto again, and the discussion quickly moves
to "gamut plots" and what not. It just broke my heart that the two of you
were sitting there just "talking" this through. No evidences, no live demos,
no worked out numerical examples for the viewer to form his own opinion,
and, especially, no actual printed sample sessions that could prove, beyond
the shadow of a doubt, what the consequences of using one color space over
the other would be. There should be a follow-up to this interview, come to
think of it. Some kind of lab exercise to follow. With a measurement
session.
The interviewer then goes on to ask "why would I want to work out in a color
space that is smaller than what my printer could reproduce"? Et voilà, the
$1000 question! How could you refute this seemingly unrefutable logic! And
your answer was "to preserve smooth transitions". Makes perfect sense to
anyone watching the video, including the interviewer, judging by the
reaction on his face, but it would help to have some worked out examples.
Next, I was surprised to hear the interviewer refering to Perceptual as the
most common wisdom RI for color. I wouldn't have thought that this was the
case. The explanation you gave over the difference of Rel compared to Perc
is "classical". The business of gamut mapping, shrinking, compression. It
seems we've all heard that too often. I was looking for something a little
more insightful...
Overall, I liked the interview. Very relax, laidback. Where was it shot?
I like when you said : "the instruments won't make the color absolutely
accurate to our color model but will keep it constant".
I'll leave my comments at that, for tonight, for I have not listened to the
whole interview.
> I mentioned it once before and there was a great silence about it on the
> forum. If you agree with it, say so. If you think it is a bunch of
> bunk, say so and tell us why you believe this.
>
> Cheers,
>
> Ray Maxwell
Thank's for bringing the DVD references up!
Roger Breton
_______________________________________________
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