Re: Color matching

On Mar 26, 6:11 am, Tom Lianza <tlia...@xxxxxxxxxxx> wrote:
Gernot Hoffmann wrote:
Roger Breton schrieb:

Tom,

Didn't you use to have a series of PDF on the old sequel.com site that
explained some of the math behind screen color calibration and matching?

Roger

Mr. T.L. didn't ever publish any scientific stuff in the Web,
as far as I know.
He's knowing everything, but he doesn't share the knowledge ...

Best regards --Gernot Hoffmann

Actually, Gernot, I did have a full set of pdf white papers that were on
the old sequel site. They were generated as the result of a high school
student's question to me on this forum years ago. Some of these papers
went through a lot of matrix math and showed examples how matrix math
can be used to solve real world problems. I got a lot of satisfaction
from letters that I got back from high school and some college folks for
showing real world examples, numerically worked out. There were also
some papers on Color Temperature and why it is a very bad way to
describe color. When the Sequel Website disappeared, much of that
disappeared, although I think there is a directory on an old ftp
mirror...I'll have to check. If you want them, just drop me a line at
tlianza AT xrite DOTCOM and I will try to dig them up...

What I never did publish was the mechanism used to determine a display
lut via inverse look up and how one could simultaneously derive RGB luts
from measurements of simple gray scale data on a display. That was
really good work and I did that when I first founded Sequel. The reason
for this is that it would basically require a chapter in a book to
describe the process fully. I am trying to get a display calibration
application with all that technology into Open Source, but it is proving
to be a bit of a challenge. The code has to be reformed using Open
Source Math libraries and Open Source USB libraries and I just don't
have the time to personally re write it, but if a college student or two
wanted to take on a little project, I'd be glad to help them. I try to
help educators and students whenever I can.

When I owned the company, Sequel Imaging, I put quite a bit of
information up on the site in the White papers section. Now that all
that technology is owned by Xrite, I am not free to just throw up white
papers when ever I feel like it.

Now as far as Mr. T.L. not publishing on the Web...the web is filled
with sites that present information, some correct, some incorrect. Many
are tainted by the author's prejudices and lack of knowledge ( I won't
use names here...). They are not peer reviewed and they are often
simply wrong. I don't need to stroke my ego by playing, part time in
this space. I normally just refer people to bruce lindbloom's site, for
color informationwww.brucelindbloom.com. I could never do a better
job than Bruce. That site is labor of love by someone who loves his
field. What the web doesn't need is another generalist publishing
non-peer reviewed information.

What Mr. T.L. does do is send between 10-20% of his pre-tax salary to
direct aid to one or two graduate students in the field of color science
or image science. I'm sharing my knowledge and my money, these days,
one student at a time... that's how I choose to make a difference.

Regards,
Tom Lianza- Hide quoted text -

- Show quoted text -


Hello

Tom, I appreciate your response. I shall take up on your offer to get
the mathematical papers and drop you a line.

As far as problem of matching the panels itself is concerned, i've
created the LUTs to match the luminance curves. As you mentioned,
though this seems to match the gray scale response, the colors
themselves seem to be off.

I then decided to approach this in the following way... There is a
document in Genot's site that specifies matrices for mapping from 1
color temperature to another (ciexyz29082000.pdf) . I manually
calculate the mapping at 8 locations on the curve ([32 64 96 ... 255])
and weigh these with the luminance ratios at these locations. I get
the whole LUT by interpolation.

I'm not sure of the concept or the logic behind this...but this seems
to give a better match for the panels (so far i've just been making
visual inspections to check if the panels match) ... I'm thinking of
extending this to more points so as to reduce interpolation errors in
the LUT.

Please advise regarding the conceptual correctness, or the lack of
it, of this approach. I would like to approach this problem in the
correct mathematical manner. I would really appreciate pointers to
other papers and documents that would enable me to educate myself as
well.

Thanks
SR

.

Relevant Pages

  • Re: Color matching
    ... explained some of the math behind screen color calibration and matching? ... Actually, Gernot, I did have a full set of pdf white papers that were on ... the old sequel site. ... one student at a time... ...
    (sci.engr.color)
  • Mathforge.net :: Transformative Hermeunetics, Euler, Nash, and a new home for young mathematicians
    ... Transformative Hermenuetics of Quantum Gravity'', which, after having ... The papers' subject matter looks interesting: ... Here is an opinion piece in Newsday on math teaching strategies. ... a statement recently against software patents. ...
    (sci.math)
  • Re: Fermats Last Theorem
    ... all 4 papers are only 26 or so pages, ... I looked at them as well, via the Math Forum's sci.math ... seen the math genealogy web page for him, ... editing and proofing my advisor and I did with my Dissertation, ...
    (sci.math)
  • Re: Current situation, suggestion
    ... > I have some papers at math journals now, so I'm just kind of goofing ... > mathematics because mathematics is distinguished in that either you're ... > find a math result, ... > papers on the fast track, unlike with that journal SWJPAM which sat on ...
    (sci.math)
  • Re: Implementation of Arctan(x)
    ... Thanks for the suggestion.Do you know any papers which, ... above concept of mixing both lut and cordic, and also good papers on fast ... hardware implementation of cordic. ... Karthik W ...
    (comp.dsp)