From: Greg Harp (gregharp_at_antispamdomain.not)
Date: Wed Nov 26 2003 - 16:18:14 EST
Steve,
In theory, since it appears that the 10D sensor (at least) is nearly linear,
if one had 75 perfectly aligned 60-second exposures and these were combined
into 15 sets of 5 by addition, the signal levels should be about the same as
15 300-second exposures. However, I'm not so sure that the noise levels
would be. Thus far I haven't seen anything indicating that noise over time
is linear with the Canon CMOS sensor. But maybe I've missed it.
However, in reality, if one has 75 60-second exposures, odds are that they
need to be aligned before being combined. And before they are aligned they
need to be calibrated, removing most of the effects of noise, vignetting,
etc., while they are still properly aligned with the calibration data rather
than each other. That's going to produce slightly different results for
noise, but in theory the resulting signal data should be about the same.
According to at least one person who posts here (Roger Clark, if I recall
correctly) 60 seconds is actually a noise "sweet spot" for the 10D.
So what I hope to do is take a large number (say 75) of 60-second exposures,
combine the using addition into (effectively) a series of 300-second
exposures, and then average those 300-second exposures. One nice thing is
that if, later in the process, I find that I would have preferred to have
240-second exposures, or 600-second exposures, I can make them.
I agree with you that, in my limited experience at least, I have seen that
longer exposures averaged together produces much smoother results than
shorter exposures simply added or adaptively added. I'm just trying to
build up those longer exposures to average by using shorter exposures that
are more within the limits of my mount's tracking capability and more
tolerant of outside interference (wind, aircraft passing through the FOV,
bumping into the mount, etc.).
I could be totally wrong here but I think that this is sound reasoning...
--Greg
> From: sc02492 [mailto:sc02492_.at._yahoo.com]
>
>
> Yes, you would think that it should work, but it doesn't seem to be
> equivalent in my hands. I've wondered whether it has something to do
> with how the frames are processed when combined (i.e., pure addition,
> in which case all of the photons would be represented in the final
> product but noise will be amplified, versus averaging, in which case
> you may not capture every photon present in each 1 minute frame, but
> noise will be cancelled out more effectively).
>
> It will be interesting to see your comparison. Perhaps "addition"
> might be the best way to combine very short exposures (low signal,
> but also low noise), whereas an averaging algorithm might work best
> for longer exposures (higher signal, but higher noise that benefits
> from stacking).
>
> Steve
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/N_XolB/TM
---------------------------------------------------------------------~->
Before posting an image to our group please read the space mangement guidelines:
http://groups.yahoo.com/group/digital_astro/files/Photos%20Section%20Space%20Management%20Guidelines
Please check the digital_astro FAQ at:
http://www.szykman.com/Astro/AstroDigiCamFAQ.html
Be sure to visit the digital_astro Gallery and Archive at:
http://velatron.com/dca/gallery
http://velatron.com/dca/archive
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
This archive was generated by hypermail 2.1.7 : Wed Nov 26 2003 - 16:24:10 EST