The only situation that I have heard of where sensor temperature has a significant effect on noise is astrophotography where exposures times can be several minutes (or hours). I believe that one technique used by astronomers to avoid temperature induced noise is to take a very large number of short exposure shots and then combine them - or use film.Chris R..
I haven't noticed any noise differences at all in either hot weather or cold weather, or how many shots I've taken..
I know you may not be looking for it, but maybe for other readers here, try neatimage:.
Http://www.neatimage.com/.
They have a free version that works well (i.e. gives you usable images, not watermarked ones), and it pretty much removes all the noise you want...
Wow, I just gave a similar answer in another thread..
There are two types of noise. The first type isn't affected by sensor temperature, the second type is..
The first type of noise has to do with signal to noise ratio. At high ISOs, you have worse signal to noise than low ISOs. This noise is apparent in dark areas. To see this noise, setup a shot at 1/100 at ISO 100, and another at 1/1600 at ISO 1600. The greater noise in the dark areas of the 1600 exposure is from this noise. If you took 16 different shots of the ISO 1600 shot, and averaged them together, it should look like the ISO 100 shot..
The second type of noise is from 'sensor build up', for lack of a better term. It happens in long exposures and it creates red, blue and white dots on the picture (the white dots my be greenish, as they are from the green pixels on the sensor). This noise appears in long exposures as some pixels build up a charge (again, for lack of a better term), and it is worse at higher temperatures. The fact that individual pixels are red, green or blue is why those colors appear as noise. A red dot is from a red pixel 'building up', a whitish one is from green pixels, and so forth..
The second type of noise is sort of predictable, as your sensor should create about same noise pattern for a X-second shot at Y-temperature degrees. For this reason, if you average 2 or more long exposure shots together, you will NOT remove this noise, as it is about equal in every shot..
The 'long-exposure noise reduction' option the camera provides attempts to remove the second type of noise, and it may be useful if you see this type of noise in shots from 1 to 30 seconds. Longer than that, it becomes a pain in the butt (as if it wasn't already)...
A lot of P&S cameras boost the Saturation, Contrast and Sharpness. Cut down these settings to the most neutral levels and you may get a stop or more of high ISO performance...
I've not noticed a big change in noise with small temperature changes..
I believe temperatures must be fairly low (like liquid Nitrogen) before you'll see much effect. Low temperatures can be achieved electronically but there would be many practical problems involving atmospheric humidity causing frost & condensation. Solving these problems would be expensive..
Here's a graph of ccd dark current vs temperature from http://www.advancedimagingpro.com/...al-Noise-in-Industrial-Imaging-Cameras/1$2642.
Image control:Zoom outZoom 100%Zoom inExpand AllOpen in new window.
The article therein says:.
"Cooling a CCD from room temperature to -25C reduces dark current by more than 100 times. Dark current doubles in most image sensors every 6C or 7C. For example, changing the temperature from 42C to 36C on a sensor with a dark current doubling rate of 6C will cut the dark current in half. If it's temperature drops from 42C to 18C, dark current reduces by a factor of 16x.".
See: http://astro.corlan.net/gcx/html/node10.html for a good discussion of noise sources...