From my own experience with old-technology CCD's for use in faint galaxy
observations with large telescopes, cooling your sensor does wonders in
greatly reducing your dark current noise. I've thought for a couple years
now that someone needs to market a thermoelectrically cooled digital sensor.
I used to cool my old RCA CCD down to -80C, and the dark current was
phenomenally reduced as a result. You're right, you need a lot of power to
do it, but considering the amount of gear we dslr photographers are willing
to cart around with us, why not have the option of cooling the CCD for long
exposure/low light shots? In the studio, extra power is no problem. In the
field, I'm sure we could lug around an extra 5 pounds, at least for short
distances, especially when we saw the low noise results afterward. I've
always found that a surprisingly spectacular result, in the aftermath of
some new hardship-inducing technique, is MORE than enough motivation to
endure that hardship. Face it, we carry around lots of heavy lenses to take
better pictures, and that's a hardship. But who would give up all those
lenses for only one lightweight lens, knowing how limited their capabilities
would become? So tack on another battery source that weighs as much as a
single decent telephoto lens... As for noise reduction, if the CCD is cooled
and maintained at a relatively fixed temperature, one could take "dark
frames" LATER, at that same temperature, and post-dark-subtract the images,
right? Meaning you could also eliminate taking dark-frame "noise reduction"
exposures during your shoot. Tell me THAT wouldn't save some time when
you're taking multi-minute exposures! The other thing I used to cool CCD's
with was liquid nitrogen, but that's pretty much out of the question for a
dslr, even though it got those temperatures WAY down there... The only
thing you have to do is thermally isolate the sensor -- whatever device is
cooling the sensor must be attached via thermally conductive "struts" to the
sensor holder itself. And the sensor and its holder must be thermally
insulated from all OTHER parts of the camera, as well as possible.
Scott
>>> Since they are stuck with 4/3rds, I'd like this:
>>> A 6 meg camera (make it the E-3a) with as low a residual noise as
[quoted text clipped - 19 lines]
> alternative would be an integrated fan and that wouldn't work too well
> on its own.
Martijn van Duijn - 12 Oct 2006 18:35 GMT
Our microscope camera is cooled to -80C. Yes, it works well, but besides
the big TEC and fan, isolating the chip in this camera involves putting
it in a deep vacuum environment. Any parts exposed to air getting that
cold result in lots of condensation/frost, undesirable near expensive
electronics!
I got the impression that building a well isolated sensor mount that is
compatible with deep cooling is no small feat (hence $$).
But also: cooling less may already make a big difference. I think the
difference between ambient and -20 is a lot more than between -20 and -80.
Martijn
> From my own experience with old-technology CCD's for use in faint galaxy
> observations with large telescopes, cooling your sensor does wonders in
[quoted text clipped - 48 lines]
>> alternative would be an integrated fan and that wouldn't work too well
>> on its own.