Return to Home Page
Hello. Please note that I changed the subject line.
> I have just visited the page and I have some questions. I have a Nikon
>Coolpix 4500 and I would like to know if I can use it to estimate some VS
>magnitudes. Using it with the telescope I can see fainter objects than using
I was interested in this topic too. But I didn't know enough
about electronics to make it plausible. Here are some thoughts
on how I reached that point. I may be totally wrong though; and
if so, I'd like to be corrected.
0. Are the dwell of these CCDs deep enough to avoid saturation?
They probably saturate too quick. Once saturates, even CCD
behaves in a non-linear manner.
1. A CCD for a typical digital camera is not the same as the professional
quality CCD. First of all, there is no embedded color filter device in
the professional CCD camera. So there is a need for three different
CCD exposures with different color filters (U,B,V,R, or I -- you pick)
to make a "true" color picture. But a typical CCD digital camera has to
capture all the colors simultaneously onto one CCD. This is accomplished
with a thin layer of color dye that covers the surface of CCD pixels, i.e.,
each pixel has a color filter to block unwanted colors. This is commonly
refered as color filtered array (CFA, no, not CfA) on CCD.
In some digital camera, the color filter array may consist of three
colors: Green, red and blue (called GRGB CFA). For those camera,
there are actually more green-colored pixels than the blue or red
ones (usually about Green:Blue:RED = 2:1:1). This is to mimic the
light response of human eyes, which are more sensitive to green light.
But this isn't the case for most of (if not all) Nikon Coolpix series.
Some Canon and Nikon Coolpix uses FOUR color filter array: Cyan,
Green, Magenta, and Yellow (CYGM CFA). Although I haven't verified,
it is generally considered that the number of pixels using each of
these colors are identical, i.e., C:Y:G:M = 1:1:1:1. So it is like
you have four color filters in one CCD camera. [This is one of the
reasons why I myself own a Nikon Coolpix camera.]
So you need to find out the response of each CYGM color filter if
you want to do really precise photometry. And you do want to know
the distribution pattern of these color filter elements.
It would be really cool if someone can hack the camera so that
it reads off raw counts per color off the camera.
[Obviously the light sensitivity of human eyes are neither G:B:R=2:1:1
or C:Y:G:M=1:1:1:1. In order to create a "true" color image, on-board
computer chip will have to do some processing to make color images.
And some fancy, professional digital camera may actually have multi-CCD
units to cover several colors at once. These are, of course, as expensive
as one of those SBIC CCDs for good reasons.]
2. Is the flux strictly conserved? The answer is not clear, as far as I
could find. First of all, everyone should save a dataset into
TIFF format. It means that the raw data are still coverted to
RGB tri-color, but at least each color layers are retrievable and
NOT flattened into one blended color image like other formats (e.g.JPEG).
Anyway, Nikon Coolpix process the detected count rates and convert
it to, I believe, 2 byte integer scale (i.e., the color intensity
is scaled into 0 -- 255 steps). So the original information on
count rates is being lost through the processing. That's at least
true for my Nikon Coolpix 995. I hadn't bothered reading up the
bloody manual to check the way around, though. [is there a digital
camera out there that let you save the data into a raw count frame?]
So having unable to do strict photon counting, you cannot do any
sort of "absolute" photometry. But if you capture your star plus
standard star in the same frame, you may be able to do relative
comparison of brightness. However note that, again, when the counts
are converted to 2-byte integer, some information is lost.
But if you can beat all these 3 points above, it'd be a nifty
thing. You can take bias and dark current images for processing
too. Or maybe I am thinking too much about details. You might
be able to circumbent some difficulty by some rigorous calibration
etc. That I don't know for sure. In any case, one needs to hack
the CCD to read off the raw counts, I'd think. Or else it'd be
just a good camera for astrophotography. But that's not what
you're thinking of doing here, right? Anyway, you'd need to ask
someone with good EE-background if you want to try.
Good luck,
Bish Ishibashi
PS. I'd think that someone has tried something like that already.
How about doing some serious GOOGLE search or something?
--
"Bish" K. Ishibashi, Ph.D.
Massachusetts Institute of Technology
Center for Space Research e bish@space.mit.edu
77 Massachusetts Ave. NE80-6011 w 617 452 5122
Cambridge, MA 02139 f 617 253 8084
Return to the Powerful Daisaku Nogami
vsnet-adm@kusastro.kyoto-u.ac.jp
