Re: ISO 12647-7 "tolerances"
Re: ISO 12647-7 "tolerances"
- Subject: Re: ISO 12647-7 "tolerances"
- From: Mark McCormick-Goodhart <email@hidden>
- Date: Wed, 23 Apr 2008 23:58:45 -0400
My research focuses primarily on photographic print permanence. The
specific requirements for contract proofs and spot color matching are
definitely out of my area of expertise, but I hope it may serve a
useful purpose in this discussion to note some basic limitations of
color difference models when they are applied to complex image color
and tone reproduction rather than to side-by-side solid color matching
tasks. For photo realistic images, there is a fundamental distinction
between side-by-side noticeable color differences and the viewer's
overall perception of acceptable image color and tonal accuracy, so
much so that color difference models fall very short in the chroma
weighting that should be assigned to low chroma colors versus high
chroma colors. HIgh chroma colors have very strong "signal amplitude"
that can tolerate much larger delta E variations compared to low
chroma colors (i.e., neutrals and near neutrals) before the color
"signal" is lost and color accuracy is judged to be poor. The ISO
tolerances being discussed here partly address this issue but not
nearly to the extent that is relevant to actual photographic print
reproduction. 6 delta E variations in high chroma colors usually do
not affect the viewer's judgement of the original scene content or
image color fidelity. If they did the visual differences between a
matt and glossy printed version of the same image could never be
tolerated. Yet 6 delta E errors in low chroma colors can throw a
distinctly perceived color cast into any print reproduction.
Likewise, delta E variations don't track image contrast. For example,
If two neighboring colors that differ by 5L lose their 5 delta L
separation, then a serious flat spot in image contrast occurs, whereas
a 5 delta L error induced by an increase from 5 to 10 L separation
causes a 2x local gamma increase in image contrast. The visual
appearance of the image in these two situations will radically differ
even though the delta E errors are the same in both instances
(assuming a* and b* contributions remain the same). I think Delta E
(pick your preferred delta E flavor) is a great process control tool,
but it needs to be put into a larger context of retained "information
content" when considering overall photographic image reproduction.
There are a couple of papers about the I* metric that address these
issues in more detail. You can find them here:
http://www.aardenburg-imaging.com/pdfs/AaI_2007_0207_TA-01.pdf
http://www.aardenburg-imaging.com/pdfs/AaI_2007_0211_TA-01.pdf
The I* metric was developed as a better mousetrap for image permanence
testing. I have always suspected it may ultimately find usefulness in
the graphic arts and printing industry as well. I'd be interested in
feedback about this metric from those of you who are more closely
involved with printing industry issues than I am.
best regards,
Mark McCormick-Goodhart
http://www.aardenburg-imaging.com
On Apr 23, 2008, at 3:02 PM, email@hidden
wrote:
Message: 1
Date: Tue, 22 Apr 2008 15:33:08 -0400
From: Kevin Muldoon <email@hidden>
Subject: Re: Colorsync-users Digest, Vol 5, Issue 134
To: email@hidden
Message-ID: <email@hidden>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Delta E max of 6??!! Delta E Primaries 5??!!
If ISO made rules for baseball then a pitchers strike zone would be
any ball that lands within 10ft from home plate! With tolerances like
that, who needs standards? This is absolutely astounding.
On Apr 22, 2008, at 3:04 PM, email@hidden
wrote:
delta E paper: 3
delta E average: 3
delta E max: 6
delta E primaries: 5
delta H primaries: 2.5
delta H avg. CMY gray: 1.5
--
Kevin Muldoon, Owner
TrueBlueDot - Fine Art Printing
New Haven, CT 06511
email@hidden
www.truebluedot.com
"Our pigment meets your imagination"
_______________________________________________
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