Dear All
I have developed a CCD spectrometer. But I can't understand how to add auto-optimization feature for best integration time. I aim is to find a integration time which is neither too high to saturate and distort the peaks, nor too low to not being able to use the dynamic range of the ccd which will make noise% more.
Thanks
I would suggest to take multiple pictures at short time scales and then add these pictures to make a high dynamic range picture. In this way, you are not limited by the bit-depth of your camera.
There is a small draw back in this technique. The noise term comes from (i) the noise in your pictures ; (ii) the electronic noise of reading the frames. If the electronic noise is small compared to intrinsic noise, the method performs well.
A more sophisticated technique is to do bracketing : take multiple pictures with increasing exposure time (usually exponentially growing). Then you can recover all your information from these pictures : bright spots from low exposures, dark spots from long exposures one.
Asmall drawback in these techniques is the exposure time, which is a bit longer than if you take only one long exposure picture.
I suggest to use a mathematical pacjkage such as matlab or the open source Julia to investigate the pictures.
Wow! Totally new perspective.
Thank you Dr. Bahram for your suggestions. Both these methods sounds exciting. I had similar doubts regarding noise addition in the first one, but I'm gonna try both anyway. I will post the results here for you to take a look. Exposure time length should not be a major issue.
Thanks again. :)
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics