HomeRoast Digest


Topic: Tom's color web page (12 msgs / 442 lines)
1) From: Robert Cantor
Tom, the green beans are brown.  If you can get a hold of a laptop or other
LCD screen you could make more accurate colors.  LCD's don't go bad the way
phosphors do.  Just be sure it can display at least 65,000 colors.  Some
only go to 256 colors.
Bob C.
rcantor

2) From: Ryuji Suzuki -- JF7WEX
From: "Robert Cantor" 
Subject: +Tom's color web page
Date: Fri, 20 Jul 2001 00:15:27 -0500
<Snip>
I don't think most CRT monitors in use today suffer from that problem.
LCDs are not perfect either. Calibrating CRT/LCD, video interface and
the viewing program is not that easy...
I saw the pictures. They seem to have the same problem with
reflection...  but it's not a major problem for many people I think...
Also, the background looks bluish. Perhaps the brightner in the paper
was excited by the illuminating light (whether sunlight, xenon flash
tube or whatever).
I'll come back to construct a rough color tile soon.
(but I won't take care of your screen calibration issue)
--
Ryuji Suzuki
"I can't believe I'm here.
People always say that I'm a long way from normal."
(Bob Dylan, Normal, Illinois, 13 February 1999)
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

3) From: Steve D - Kc4rkf

4) From: Steve
I don't see any problem. They look pretty darn close to the color I usually
see while roasting. I suppose we could get into color blindness, color
variations by race, moisture, external temperature and lighting, length of
time staring at the pics...
how we spend our time...
Tom, excellent samples but glad to see you spending your time finding the
best beans.
steve
Perfect in Seattle

5) From: Irene and Lubos Palounek
I fully agree with Steve:
"Tom, excellent samples but glad to see you spending your time finding the
best beans."
However, the colors look quite differently on our two computer monitors. I
think it would help:
 -- use the "Kodak 18% Neutral Test Card" as the background.  Anybody can
buy that card inexpensively in any good photo store.
 -- use the "Macbeth color chart" as one of the pictures.
(Macbeth is now allied with a prominent Swiss company, Gretag. The new
combined company is GretagMacbeth - no space between the words. Seehttp://www.gretagmacbeth.comfor more information.)
I would also like to see the colors of ground beans, similar to the
Agtron/SCAA charts.
BTW, how stable is the color of ground beans?  Would it make sense to
provide samples of the ground beans roasted from the #95 to #25? Or perhaps
fill a set of Petri Dishes with the samples and sell the set?
Thank you again, Tom, for the great help you are giving all of us.
Cheers, Lubos
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

6) From: Steve
Creating Web sites has always been a problem due to the wide variety of
monitors, video cards, and, most important, what each users video settings
are. You could create the most perfect image but it will still come out
differently. There is no way to control what everyone sees.
If you want to get the best view you should have a video card of 8mb or more
(the new ones are way past that now), your settings are at the highest
(usually 24 bit or higher) and a monitor that can handle the hightest
settings, you should see the correct colors.
steve
in Seattle

7) From: Ryuji Suzuki -- JF7WEX
From: "Steve" 
Subject: Re: +Tom's color web page
Date: Thu, 19 Jul 2001 23:53:28 -0700
<Snip>
To Steve in Seattle. Unfortunately, you are talking about different
thing.  I'm not talking about color. Image reproduction is extensively
studied in quantified mannar and I'm glad that there is not much room
for shaky arguments, compared to taste problem, which is much less
studied and one can always run into subjective issues, personal
preference, etc.  As long as your monitor and graphic interface are
properly calibrated the color problem is minimal (although not
eliminated), and I need not spend my time to discuss about your
argument.
I think inclusion of Macbeth ColorChecker, which Lubos suggested is
indeed a good idea and helpful in interpreting the color of the
roasted coffee in the image, but I guess most people who have nothing
to do with photography or printing are not used to interpreting the
reproduced chart... (Note: Macbeth chart consists of 4x6 boxes filled
with color - bottom row is 6 steps of greyscale, from row 1 to row 3,
each row has three additive and three subtractive primaries. This is
NOT made by a CMYK or any four-color printing, but printed with
carefully formulated pigments to make accurate reproduction of the
chart as challenging as real objects or scene.)
--
Ryuji Suzuki
"I can't believe I'm here.
People always say that I'm a long way from normal."
(Bob Dylan, Normal, Illinois, 13 February 1999)
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

8) From: Robert Cantor
Then you'll be able to sell it.
Bob C.
rcantor

9) From: Henry C. Davis
Only if some patent lawyer doesn't get there first!

10) From: Tom & Maria
I did notice that "yellow" and "brown" images appear to be switched, and
one needs a bit of color correction. In terms of color accuarcy... I was
shooting for representations of these roast colors so slight variation isnt
going to make  the evaluation of the roast images null. And yes, I should
add ground samples too ...I forgot!
Tom
<Snip>
                  "Great coffee comes from tiny roasters"
           Sweet Maria's Home Coffee Roasting  -  Tom & Maria
                     http://lists.sweetmarias.com/mailman/listinfo/homeroast">http://www.sweetmarias.comhomeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

11) From: Irene and Lubos Palounek
For some reason, the following item I sent last Saturday shortly after noon
never made it to the list.  So I am sending it again.
Cheers, Lubos

12) From: Ryuji Suzuki -- JF7WEX
Although it's getting a bit too technical...
(it happens anyway)
Those who are not interested in image reproduction issues, please skip.
From: "Irene and Lubos Palounek" 
Subject: +Tom's color web page
Date: Mon, 23 Jul 2001 23:25:35 -0500
I agreed that Macbeth Color Checker is a good idea, although it costs
a bit.
<Snip>
The color temperature of the lighting should be matched to the
spectral sensitivity characteristics of the emulsion (in case of
silver halide photography) or CCD or other image sensors (in case of
e-imaging).
<Snip>
Color matching is relatively easy, but it's not the end of the story!
Contrast curve, or at least gamma value must be also matched for
accurate reproduction. (Without this calibration, the image appears
like B&W negatives printed on wrong paper grade!)  I have been trying
to make a web page to calibrate contrast so that people can see my
photographs with best possible setting, but I have been spending more
time in darkroom than in writing web pages. (WWW stands for World Wide
Waste-of-time)
<Snip>
I don't know the detail of that technology, but my guess is that it is
just another marketing hype. For example, even if there is a way to
match R, G, B curves automatically to a reasonable tolerance, it won't
take care of color saturation problem, etc. For example, if there are
vivid colors and pastel colors in the same image, it is practically
impossible for both groups of colors to be accurately reproduced with
faithful saturation level. That gives a big motivation for
photographers and printers to test films and papers to find pairs they
like for their images. (and I have my preference, although I rarely do
colors)
A typical computer-head thinks that video memory megabytes and number
of bits are all they need to think about, but this is by no means the
case!!!  (right, how accurate is good enough for the job is another
issue, and most people who don't have anything to do with photography,
e-imaging, printing and publishing don't need the kind of accuracy
that I am talking about here. It's just like SM's numerical rating
system is not good enough for us and it's too much for insensitive
coffee drinkers.)
<Snip>
Human eye has different response to various light sources, compared to
CCD or films. Because it adaps to lighting condition fairly well, I do
not think that high degree of color temperature accuracy is necessary
for judging roasted coffee. However, using the same lighting and grey
or cream background material (avoid pure white or black) each time is
likely to be very helpful in calibrating the eyes to the light
source's color balance.
<Snip>
The best way is to use two polarizers, one on light source and another
on taking lens. (this is necessary if the object has metallic reflection)
--
Ryuji Suzuki
"I can't believe I'm here.
People always say that I'm a long way from normal."
(Bob Dylan, Normal, Illinois, 13 February 1999)
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast


HomeRoast Digest