Quote:
Originally Posted by JohnH
Thanks for that advice Doug - it is a subject I have not been able to get a straight answer on before so I have used trial and error till now...if I may I have a couple of questions...I think I get the offset routine but for GAIN I am not sure.
With my CCD I can see a relationship between GAIN and NOISE - thus more gain means more (thermal) noise for a given exposure time. Now the signal and noise rates do not increase at the same rate with gain meaning a gain increase may double signal strength but only increase noise, say, 10% thus you have better S/N ratio which is good. Go too far and the S/N ration will worsen.
So S/N ratio may set an upper limit on gain.
If it does not (you have little or no thermal noise) then increasing GAIN too far will eventually lead you to loss of bandwidth as more and more of your image is saturated (clipping).
So what is the best gain setting to use? If I try to use your suggestion and use my camera in daylight it will saturate in no time - for example to expose my flats I used an exposure of 0.05s - this gave a max signal of 16,928 average 15,158 and min of 128, in a 4 min star shot I have a max value of 65,520 but average 15,120 and min of 6,864.
So I think my offset is ok but I am not sure about the gain. Based on visual inspection of the attached charts I think I am best off in the 200-400 range.
Any comments?
Frank sorry to divert the thread....
It is a great first LRGB I am on this learning curve with a 127mm also (not the same CCD though). Tell me did you find any significant differences in the star sizes between RG and B frames? I find my B images are soft compared to the others it may be my filters though - I use the Astronomik LRGB set.
|
The gain settings etc should not be changed relating to noise etc.
The AD converter on you camera is 16 big meaning that you can output 0 -65000 (aprox) different levels per pixel. I don't know the full well capacity for the CCD but it is measured in electrons. The gain is simply the number of electrons that are added together to get an output by the camera of a 0-65000 level to read into your picture.
If the full well capacity is 100000 then a gain of ~1.3 will use the full depth of the CCD and you get useful data in the entire range of the output.
As stated in Dougs post you need to first set an offset in the dark so that there is a base of about 100 in the image. This is just to prevent negative pixels.
Then take exposures of anything bright trying to completely saturate the CCD with any exposure time. The gain then needs to be adjusted to so that the totally saturated exposures read about 65000. Keep reducing the gain until you start to get readings just below this value.
This tells you that you are using the full wells of the camera.
Once you know this valie don't change it.
When you image a dim object you will end up with low values but dont increase the gain. The images are easier to stretch when you have 65000 levels to play with rather than less than that.