Sponsored By:

Subject: [M]: RE: Meade 416 CCD and Color Wheel
From: Michael Hart
Reply To: mapug@shore.net
Date: Fri Mar 13 02:11:49 1998

On Monday, March 02, 1998 1:59 PM, Britt, Dale (AZ15) wrote:

> I finally received my long awaited Meade color wheel for my 416 XT. However,
> I either don't know how to find or don't have the information to determine
> the Red-Green-Blue exposure ratios for the 416 XT. Can anyone provide me with
> this information or tell me where I might find it. I will appreciate any help
> in this area.

CCD color filter exposure ratios is a very complex subject because so many
factors effect the optimal ratio including the desired results. We have come
to expect astronomical objects to be quite colorful. Often, the goal of image
processing is to produce a final image with the colors we desire rather than
the colors that represent actual wavelengths. Part of that processing starts
with selecting a background and range in the histogram because we cannot
see so many shades of color at once. In this case, we compress a rather
large dynamic range into a smaller range that we can see. Worse, computer
monitors and photographic prints have an even more limited dynamic range.

As a result, we may use some type of non-linear scaling to compress and
re-map the range of colors and brightness to the relatively narrow frequency
response that we can see. Using non-linear scaling will effect the color
distribution of the final image even if we start with a carefully calibrated
tricolor exposure ratio. I have demonstrated this in lectures- I took
emulsion prints of M-42 where I only alter the scaling. One looks pretty
much like what we see in many prints of this object. Another looks exactly
like recent CCD images of this object.

No matter what ratio you do use, it is important to obtain a good S/N
(signal to noise) ratio in EACH color or you images will appear noisy
(colored speckles) which limits ultimate enlargement potential. Careful
attention to S/N ratios before adjusting color balance has the potential
to produce final enlargements to 80-90X and beyond. I am not speaking
of projected images that have the illusion of detail do their physical size.
Rather, I'm speaking of large prints of the 16 X 20 class where you move
to within a few inches of the print to discern fine details.

Using a color filter set comes at the price of much longer blue exposures
when using the Kodak 0400 chips. In your case, I'd start with 1, 1 and 3
based on the spectral response curve of these chips and watch signal
to noise ratios of the individual colors. I suggest using 3 for the blue
because the Meade set receives considerable energy from the green.
Other sets with less crosstalk and sinusoidal bandpass require
much longer blue exposures, however the final results quite pleasing.

You can use a neutral gray card under a full moon to gauge exposure
ratios with a particular CCD chip, but one must account for atmospheric
attenuation of particular wavelengths from the zenith to around 45 degrees
where images are often made. In this case, red is least affected, followed
by green and much more so in the blue. Even if you obtain perfectly
balanced exposures at optimal atmospheric attenuation, you will
most likely alter that balance with subsequent image processing.

Tricolor filters and methods are discussed in great detail on Doc G's
web page at:

http://www.mailbag.com/users/ragreiner/index.html under
CCD Imagers and Accessories, Three Color Filters for Film
and CCD Imaging with further discussion in Color Imaging
Methods with the CCD Imager.




> A second question about the color wheel. Mine is quite noisy when it moves to
> a new filter, is this normal.
>
> I am using the new Pictor 6.4 Beta 2 and have uncovered a bug that I will
> report to Meade. Thought it might be of interest to the group. I am using the
> SCSI interface, and found that if you take a color image using a profile that
> has automatic dark frame on you get garbage for the downloaded images.
>
>


--
Michael Hart
Husen Observatory
mhart@netexpress.net