On Monday, July 06, 1998 7:14 PM, R. A. Greiner wrote:
>
>
> Bob Freeze wrote:
>
> > Well, I finally received my filters and adapters. While waiting for
> > clear sky I decided to see if I could figure the proper exposure
> > ratios for my particular camera/filter combination.
> >
> > In order to figure this out, I set the 416XT up in the darkroom so I
> > could use the enlarger as a light source. BIG ASSUMPTION - the
> > enlarger produces light equally across the visual spectrum. I think
> > this is true because I have zero enlarger color factors when doing
> > color prints.
> >
> > Anyway, I found that the following exposures produced equal pixel
> > values:
> >
> > RED (25) = 1
> > GRN (56) = 1.62 (yes, I know that 58 is the standard GRN for this
> > set)
> > BLU (47) = 3.5
> >
> > With a RED exposure of 1 second (and GRN/BLU set accordingly), I had
> > peak values in the same part of the image of 47,000 as reported by
> > PVXT 6.4.
> >
> > If this is accurate, I'll be looking at just over an hour for a set
> > with a 10 minute RED exposure!!
> >
> > Bob
> > http://www.usit.com/rfreeze
> > (~36N/84W)
>
> This is very interesting and solid data which I believe to be correct.
> There is one major ussue you need to take into account. The color
> temperature of the enlarger bulb is probably about 2800 to 3200 Kelvin.
> The color of a star is more like 5800 to 6000 Kelvin. Thus the star
> which you might like to have look normal (neutral or white) will have
> much more blue in it than the enlarger light..
>
> I would expect your exposures to be somewhat more balanced in time than
> you anticipate from your data. Never-the-less, the blue exposures are
> always rather long because of the low sensitivity of the chip to blue
> light.
>
> There is some rather tetailed discussion of these issues on my web site
> under. Attachments/Filters which you might find interesting.
>
> Doc G <http://www.mailbag.com/users/ragreiner>
>


Doc G is correct about the desired color temperature needed
to best calibrate RGB filters. The 5800 to 6000 K corresponds
roughly to a G-2 type star positioned with minimal atmospheric
attenuation of the shorter blue and to a lessor extent, the green
wavelengths. Why 5800 degrees K? Because this corresponds
to the effective temperature of the sun (5770 degrees K., spectral
type G2) which the human uses to see color.

Why bother with calibrating color filters if one can adjust the color
balance in the software? Because achieving relatively balanced
exposures also balances the S/N ratio of each channel. As a
result, greater color saturations (and more) may be considered due
to the minimized noise. With stacked images (shift and combine or
track and accumulate), poor calibration times may result in
unexpected colors.

There are relatively few bright G2 stars brighter than Mag 3 (only
three come to mind) I believe the best and easiest way to calibrate
a selected filter set with a particular CCD camera is to use the sun
itself. I use a full moon quite high and a neutral gray test card,
a camera store item. John Hoot has suggested a way that may
be simpler- image the moon directly. However, I would suggest
stopping down the telescope, if needed, rather than using a neutral
density filter which may not really attenuate all wavelengths equally.

Are exact ratios critical? No, because atmospheric attenuation
that varies with altitude and atmosphere conditions will cause the
ratios to vary quite a bit. If you expect to image quite low, wait
until the moon is near the altitude you will be imaging and calibrate
your filter set. Don't forget to use an IR filter with Wratten and dicroic
filters to assure IR leakage doesn't skew your results.

How can one determine the initial ratio? Start out by comparing the
CCD chip's QE (quantum efficiency) in the center of the wavelengths
passed by your filter set. With Kodak 0400/1600 chips, a filter set
that results in surprisingly low blue exposures is likely leaking light
from non-blue sources such as infrared and green and possibly the
red channels.



--
Michael Hart
Husen Observatory

View index by [date] [author] [subject]
Previous message: [M]: Database Version, Steve Webb
Next message: [M]: Image processing software, Gary McKenzie
Next message in thread: RE: [M]: 416XT RGB exposures, Heapy Chris CG
Previous message in thread: Re: [M]: 416XT RGB exposures, R. A. Greiner