PDA

View Full Version : Hot pixels



keystar
November 19, 2007, 20:04:52
Hi Stefan,

U wrote this:
"I understan, waiting an hour for an image is not as funny :-) We tested this. The image is not as noisy. I suggest: Close the camera (as it as delivered) and start to expose for one hour. Go and drink some coffee. Then you may handle your shoppings :-) come back and see the image. I would use the automatic image sequence and set it to a 0 time intervall. Then all incoming images are saved."

My problem is:
1. DBK 41AF02.AS
2. Ambiant room temperature
3. camera close (as it as delivered)
4. EXPOSE ONLY 32 seconds

See picture below:

question: what's wrong???
thx

Have a nice day!!

Stefan Geissler
November 20, 2007, 09:03:38
A pixel is "dead" if it is stuck to a certain graylevel (often "white" or "red", "green" and "blue" in case of color cameras) independently of the gain and/or exposure time settings. These pixels are easy to detect and thus CCDs with a dead pixel do not pass our QA. However, it is possible (but very rarely) that a hot pixel appears during the lifetime of a camera, and... obviously... even our QA may fail sometimes ;-) .

In contrast, *every* CCD has hot pixels. In theory all pixel should behave equally but in practice they don't. You do not see these slight differences in a normal operation mode but if you increase the gain and/or exposure time drastically (leaving the CCD protection cap on the camera to avoid any effects due to normal light) you get images as the one I attached to my last mail. In this case the exposure time was 2s and the gain was at its max.

We exchange cameras with dead pixels but obviously we cannot exchange cameras with hot pixels.

In you case the gain may is set to automatic or to a higher level. This creates an extra portion of noise.

oldfrankland
November 21, 2007, 15:56:54
Hi Stefan,

U wrote this:
"I understan, waiting an hour for an image is not as funny :-) We tested this. The image is not as noisy. I suggest: Close the camera (as it as delivered) and start to expose for one hour. Go and drink some coffee. Then you may handle your shoppings :-) come back and see the image. I would use the automatic image sequence and set it to a 0 time intervall. Then all incoming images are saved."

My problem is:
1. DBK 41AF02.AS
2. Ambiant room temperature
3. camera close (as it as delivered)
4. EXPOSE ONLY 32 seconds

See picture below:

question: what's wrong???
thx

Have a nice day!!

What you're seeing is typical of all CCDs when used for extended time exposures. The hot and warm pixels will be particularly prominent at room temperature. Do the same test outdoors on a cold night and the visible pixels will diminish considerably. If you intend to do long exposure deep sky astrophotography with the camera, stacking a number of sub frames and subtracting a dark frame will completely do away with the effect of hot and warm pixels. Standard operating procedure.

If you are using the camera at shorter exposures for lunar and planetary, there is no issue at all with hot and warm pixels. With one exception: if your mount tracks perfectly, where the same pixel is in exactly the same place in the object for each sub frame image, you could see bright or dark spots due to hot or warm pixels. Kind'a rare, I should think, but possible, though I suspect seeing conditions would still smooth out the noise.

Cheers,
Jim
Livermore Ca
http://www.lafterhall.com/ccd_imaging.html