Hi, thanks for your reply.Thank you for sharing these results, quite interesting. One thing to note that these values are given in 16-bits, but the sensor works in 12-bits and the final ISP processed output is 8-bits. So the actual difference in dark current (black level) would be quite small in the output. I did have a couple of questions:
1) For your first graph, is the analogue gain fixed to 1.0?
2) I wasn't quite sure what your second graph was showing - can you explain the value on the y-axis?
I agree wrt to the 16-bit values. To capture this I am using rpicam-ascom-alpaca which is then connected to N.I.N.A.. I beleive rpicam-ascom-alpaca is stretching the output to 16 bits but it is likely doing a linear stretch. I have raised an issue about the bit depth here.
Please can you explain a little more about the 8-bit output you mentioned? I thought it was possible to get the full 12 bits out as a digital negative, which is almost certainly what rpicam-ascom-alpaca does?
By the way, I tried rerunning the experiment in the freezer where the dark current would be reduced, and I saw a similar pattern but now the lines overlap so I'd struggle to account for that for a given exposure time:
To answer your questions:
1) I beleive the gain is the default setting as I haven't specified it and in N.I.N.A. it's set to default.
2) This is a plot of average pixel intensity versus exposure time. The camera was set up facing a torch pointed at it through a diffuser to give a uniform light field. I wanted to test up to what pixel values the response from the sensor is linear (as this is the region you'd use for astrophotography). You can see it stops being linear at about 50,000. You'd expect the linear trend to continue down to the y-intercept, but I see a strange plateu. I wondered whether this was down to extremely low exposure times, so I reduced the aperture for the camera and did it again, and you see a very similar result. So this leads me to beleive it's something to do with how the sensor works.
I have posted about this on the Cloudy Nights forum:
Dark Current
Exposure Plateu
Statistics: Posted by DanielDC88 — Fri Dec 13, 2024 1:55 pm