Re: color calibration of microscopic images

• Previous message: Aaron: "Re: color calibration of microscopic images"
• In reply to: Aaron: "Re: color calibration of microscopic images"
• Next in thread: Gordon Couger: "Re: color calibration of microscopic images"
• Messages sorted by: [ date ] [ thread ]

Date: Thu, 2 Sep 2004 15:39:26 +0100

In article <oiibj0h0s2u8ukus51jc550dktdmvh9b6q@4ax.com>, Aaron
<nghy@comcast.net> writes

Nice response, Aaron - I told B she would get much more help here than
in rec.photo.digital!

David

>As a professional chemist, I worked on color matching problems.
>Digital photomicrogrphy is a current major interest. I have had
>occaion to work with many of the digital cameras both consumer and
>professional dedicated microscope cameras including the Nikon that you
>are using..
>
>Before I address some of the technical issues in measureing coloit,,
>I'd like to point out that artists mix pigments quite freely while
>creating a painting, and it is probably the case that your samples
>will not be homgeneous nor will they be consistant over different
>portions of the canvas. You also need to review the processing of the
>microscopic color specimen because the method of abrasion and the
>abrasive itself will affect the surface texture and may also leave a
>trace residue embedded in the specimen surface which may also affect
>color.. .
>
>Even bypassing the issues of sampling and sample preparation there are
>equipment issues and human perception issue which complicate your
>work.,
>
>I think you will benefit from additional study of human color
>perception. In short, colors, as perceived by the human eye/ brain
>combination, can be represented as existing within a three dimentional
>solid. Each color is a point within the solid. There are two
>coordinate systems which may be used to specify the color; X,Y,Z or
>spherical. . The vertical axis of the solid is brightness and goes
>from black to white. The hue and saturation are defined by the
>direction and distance from the origin. Both systems accomplish the
>same objective although the nomenclature is slightly different..
>
> Humans may perceive the same color source as being several different
>shades depending on the lighting conditions, the orientation of the
>light source,, surface texture and translucency. Humans may
>perceive two colors as being the same even when the reflected spectra
>are substanially different. In reverse humans may see differences in
>color when the reflected spectra are nearly the same.. The reason for
>this is that the mind processes the data received by the eye and makes
>judgements which we perceive as color. Any experienced photographer
>knows that a camera sees very differently than the eye/brain
>combination.
>
>Part of your problem is an asumption that there is an absolute color.
>Colors occur with specific conditions. Color matching is a function
>of the test conditions. So it is normal that the observations made
>with the microscope will be specific to that set of conditions and
>that you will find it very difficult to duplicate those colors under
>other conditions.. .
>
>There are colorimeters that use tristimulus values obtained with three
>broad band filters, i.e. the Hunter Colorimeter. These will provide
>coordinates within the color solid and be analogous to that humans
>perceive.. There are also spectrophotometers which can provide a
>measure of the reflected light as a function of the wavelength of the
>light source., which can provide clues to the pigments involved. The
>spectral curves are good scientific data but are not directly related
>to human perception of color..
>
>I think you might need to look into using these types of instruments
>rather than a digital color camera.
>
> In all my work with photomicrography, I have never found a color
>digital cameras in any price range that provides highly reproducable
>or accurate color renditions. The cameras will come close enough in
>rendering the colors for us to recognize the objects that have been
>imaged, but that is it.
>
>In addition to the camera, there is a chain of devices which convert
>the digital data to viewable images. Each of these devices
>introduces its own effects on the colors, i.e. the graphics card in
>the computer, the monitor, and/or the printer.all modify the colors.
>You cannot assume the digital information will provide similar
>renditions of color on different equipment. And even over time the
>same equipment will change its charateristics. There has been work
>done to minimize gross errors in these systems through the application
>of special standards and color management software. In my own work I
>find photoediting software indespensible in correcting the colors,
>sharpness etc.. .
>
>You may find the basics covered at this website sponsored by Kodak.of
>interest. I appologize in advance if it is too basic.
>http://www.kodak.com/US/en/digital/dlc/book3/chapter2/index.shtml
>
>
>There are probably more variables in your system than you may have
>recognized. It may be helpful for you to devise a procedure which
>eliminates as many as possible and generates numbers for comparisons
>over time.. For example, one such method might involve choosing one
>objective and microscope setup for all measurements, attach the output
>of the scope to a colorimeter and evaluate all specimen both large and
>small with the same setup. Of course i do not know enough about your
>application so what I have suggested is not likly to be the answer. I
>hope you get an idea from my comments.
>
>Good luck
>Aaron
>
>

-- 
David Littlewood

• Previous message: Aaron: "Re: color calibration of microscopic images"
• In reply to: Aaron: "Re: color calibration of microscopic images"
• Next in thread: Gordon Couger: "Re: color calibration of microscopic images"
• Messages sorted by: [ date ] [ thread ]