Hi all
I was recently reviewing a DVD by Adam Block, where he describes the 3 calibration frames in detail.

Adam suggests that a single master bias frame, comprising many individual bias frames, can be used to calibrate any image regardless of what temperature the bias frames were taken.

I have an STL11K and my experiments with this yield rather contradictory results. In fact, so different are the results that I have bias frames for every temperature of light frame and bin mode.

Here is what I found:

Temp ADU Values1x1 ADU Values2x2
-10 Max - 7058 Max - 3632
Avg - 739 Avg - 446
-15 Max - 4868 Max - 2436
Avg - 732 Avg - 430
-20 Max - 3713 Max - 1955
Avg - 729 Avg - 420
-25 Max - 3008 Max - 1551
Avg - 722 Avg - 408
-30 Max - 2489 Max - 1314
Avg - 720 Avg - 406
As you can see..the average values seem fairly consistent but the max values are quite high.
So.....use a -10 degree bias frame, will introduce noise into a -30 light frame.

Also, Adam suggests using just a bias frame to calibrate flats.....no way!! Not with the STL 11K any way. Even a 5 second flat frame at -5 degrees has too much dark current in it. If my target ADU for my flat is 25,000 counts, I find I have to calibrate the flat with a dark (scaled using a bias as the exposure time would vary when doing sky flats), otherwise, my resultant ADU are way off. However, if I calibrate carefully in this way, my average ADU is always around my target flat ADU.

Thoughts?

thanks
martin

Hi Martin,

Actually, your table shows that there is very little difference and a one-temp-fits-all bias is probably just fine.

Your 1x1 gain is near 0.8 so an elevation in temp of 20C only adds about 15 electrons to the bias level, which is utterly trivial for a flat, which should have an avg near 30,000 electrons.

Furthermore, if the bias is a master (made from many frames) then it contains virtually no noise (about 1-5 electrons, depending on how many frames) and the 15 electron difference is only an insignificant offset and not a source of noise.

There are a very few hot pixels that will show a significant temp variation (your "max" numbers). But they are going to malfunction anyway and be replaced via data rejection. It is foolish to obsess on a handful of worthless pixels, so don't pay much attention to "max" numbers.

Stan

Thanks for the observations there Stan...I appreciate it.

cheers
Martin