Sensor analysis help for an ASI2600-MC-Pro

[Lee_P]

Hi all,

Today I bought SharpCap Pro, with the eventual aim of using the Histogram feature to give an indication of optimal sub lengths. Toward that goal, I'm trying to run the sensor analysis. I'm not sure the results I'm getting are realistic, so would appreciate some wiser eyes than mine to take a look!

Lee's ASI2600 sensor analysis.JPG (221.05 KiB) Viewed 1136 times

I followed the instructions and have run the analysis four times, but always with similar results. I never get an e/ADU of one, which I think is odd? If so, I'm probably making a basic error someone along the way. Any suggestions?

Thanks!

-Lee

[admin]

Hi,

those results look pretty reasonable – the only thing I'd say is that it looks as though the camera is switching from LCG to HCG mode somewhere between gain 61 and 100 (where there's a big drop in the read noise). SharpCap expects this to happen at gain 64 ZWO cameras (at least that was where it happened for most of the early cameras that had the LCG/HCG switchover). Since SharpCap doesn't have any exact knowledge of where the switchover is happening, the suggestions for exposure length may be incorrect in that gain range between 61 and 100 - SharpCap will be assuming that the read noise changes linearly between those two values when in fact it will fall with a cliff edge at whatever the switchover point is. Basically, the message there is that if you are using the suggestions from the smart histogram either stick to gains below 61 or above 100.

Cheers, Robin

[Lee_P]

Thanks for the speedy reply Robin, that's a great help. I'll run the smart histogram on the next clear night using the current sensor analysis.

FYI, ZWO recommend using this camera at gain 100:

ASI2600.JPG (28.02 KiB) Viewed 1131 times

Thanks again!

-Lee

[PetePeterson]

Did my first sensor analysis on my ASI2600MC Pro this week too. Got the same result. Figured noise is bad. The lower the read noise the better signal to noise ratio, no? So the 200 gain point seemed to be the break point and that's what I've been using as a standard setting. Do I have this backwards? I may very well be interpreting this incorrectly.

[admin]

Hi,

read noise is bad (it would be great if it was completely eliminated), but by taking long enough exposures you can effectively wipe out the effects of in the final stacked image. The key thing is that the length of exposure that you need to take is proportional to the square of the read noise. That means that by moving from gain 90 where the read noise is >3 to gain 110 where its less than 1.5, you can reduce the sub-exposure length required by a factor of four. Great news if you are struggling with guiding or tracking.

If you don't mind a bit of mathematics then it's all explained in great detail here : viewtopic.php?f=35&t=456

Cheers, Robin