Thank you, Brian:
You are right, your calculations illustrate that the file size is,
indeed, quite unwieldy.
The display tools that I've seen on the web so far aren't just right,
and they cost a lot.
Graphic designers use the electronic sensor probes and specialized
software to adjust the displayed screen colors. For those purposes I
would want several application specific ranges of color for use where
color is really im****tant: for humans, for foliage, printing inks, and
for atmosphere.
Some of the web and print graphic designers color requirements are
irrelevant to celestial photography. Astronomers don't have any of those
color requirements. The requirements are so far that I can imagine them
are:
* overall screen brightness [not too bright or too dim]
* adjustable overall screen brightness and contrast [to discern stars,
celestial details, Moon and planetary images.]
* overall screen color balance [for R,G,B]
* overall displayed contrast range [an even gradation over a span of 20
to 50 or 255 steps from:
black to white [0,0,0 to 255,255,255]
* adjustable gamma and gamut controls [for enhanced or conserved images
- also see Photoshop]
* adaptations for artistic color enhancements
to bringing out certain types of image data
to make image features recognizable or to enhance color effects.
* adaptations for images that have been made with specific light source
filters
[expert information here needed]
* stacked images
[expert information here needed]
While Photoshop can do many of these things it is about as far from
WYSIWYG as you can get. Photoshop has good capabilities for image
manipulation, however it doesn't change monitor settings. A dynamic
screen image and color control system would be most of interest, so far
that I can imagine.
Astronomy image stacking and manipulation software does exist
[expert information here needed] [I know little about what the
specific software brands can provide.]
The scope of those requirements is too great for the simpler task of
monitor contrast settings.
I'm obviously involved in a learning process.
The immediate need that I see is for a color map display file that one
could bring up in any paint or image display program, e.g., Paint Shop
Pro or Irfanview.
That file would have the four graduated bands of color:
black to white [0,0,0 to [255,255,255] [for displayed contrast
evaluation]
black to red [R 0-255] [to evaluate the effect of certain changes to
monitor settings]
black to green [G 0-255] [to evaluate the effect of certain changes
to monitor settings]
black to blue [B 0-255] [to evaluate the effect of certain changes to
monitor settings]
24 or 32 bit color monitors [and hi-res display adapter cards] are
essential for the task. Those monitors can display 255^3 possible color
mixtures of RGB colors. Monitors that display only a total of 255, 24,
12, or 1 colors are not too useful for astronomy images.
The number of displayed key colors in the file would only need to be 4
times 255. Each color would have a color number in the RGB format, e.g.,
[255,255,255] for white. Grays have equal numbers in the RGB values,
e.g., [56,56,56] for a dark gray.
In time I"ll check my MicroStation CAD, or other CAD, resources to find
some color maps to make into a file. MicroStation has color tables that
I could adapt, and I would then make a drawing and apply the colors to
the key areas in the drawing. By the time I make that file I'm sure that
I will find something that is available on the market.
The displayed color file would have the appearance of:
|
[ 255 grays ]
[ 255 reds ]
[ 255 greens ]
[ 255 blues ]
|
I would make the file into an uncompressed .tif file that you suggest,
and it could be display in most image display programs and be printed on
a color printer at A or B photo paper sizes.
Maybe I'll just do the gray scale first and post that. That will have
the most value for monitor setup work. The other scales would be more
for filtered image evaluation or for color printing.
I'll make the CAD color scale drawing for grays:
|
[ 255 grays ]
|
The gray scale band would have 255 color areas. Each area could have the
size: 1cm H by .5cm W. I'll make a single band of grays. That could be
displayed at appropriate sizes and printed on all paper sizes.
Resolution doesn't matter. The .tif file will have sufficient resolution
to be printed neatly on 8.5x11" or 11x17" paper, for example.
I'll make a single band, and that may be evaluated. An option is to
break the band into four bands.
Thanks again for the information.
Ralph Hertle
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Skywise wrote:
> Since 256^3 = 16,777,216 and you need at least one pixel to
> represent each possible color, you need at least that many
> pixels, which results in quite a large image. 4096 x 4096
> would do it quite nicely.
>
> Tiff format, even compressed, would be quite large. JPG
> compression would destroy the subtle color detail and
> would be worthless.
>
[...]
|
