ASI2600 and Dark Frames

Anything that doesn't fit into any of the other forums
Post Reply
nexusjeep
Posts: 293
Joined: Thu Oct 18, 2018 3:01 pm
Location: Gloucestershire

ASI2600 and Dark Frames

#1

Post by nexusjeep »

Hi I have just purchased the ASI2600MC-PRO and have run the sensor calibration and am now in the process of considering dark libraries. Having had a play there is no discernible amp glow which is one of the features of this sensor. On my ASI1600 I have to shoot darks for each of the exposure + gain settings as there is significant edge glow from the amplifiers.

However with the ASI2600 I do not see any indication of amp glow and have examined test frames up 5 minutes so far, and even with the auto stretch in SharpCap do not see anything. So the only thing that the dark is going to record is hot or cold pixels so is it possible to shoot just at one exposure say 60s at gain 0 and gain 100 which are the two gain settings that I will use the most and then use dark scaling within APP, etc to correct these hot or cold pixels.

Anyone have any thoughts on this.

For info the sensor analysis from Sharpcap is below this was done at -10C as I should be able to hold this anytime of year.
ASI2600.png
ASI2600.png (95 KiB) Viewed 2007 times
All I need now is to actually be able to see the sky.

Cheers
Nick
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: ASI2600 and Dark Frames

#2

Post by admin »

Hi,


Personally I would still choose to make dark frames perfectly match the settings I use for my light frames. Although the camera doesn't show a significant amp glow, darks still do much more than just correct hot pixels… In particular dark frames capture the average rate at which the camera generates thermal noise signals which add to the output signal from the camera. You do want to subtract this correctly when you are processing and for that you really need dark frames to match the light frames.

An alternative, if you are confident that the dark noise in your camera is well behaved is to do as you suggest and capture a single length of dark frame and also capture bias frames. Processing software such as DSS or pixinsight can use the fixed length dark frames and the bias frames to estimate what a dark frame of a different length would look like if necessary during the image processing.

Cheers, Robin
nexusjeep
Posts: 293
Joined: Thu Oct 18, 2018 3:01 pm
Location: Gloucestershire

Re: ASI2600 and Dark Frames

#3

Post by nexusjeep »

Hi Robin,
Thanks for that, I have taken that approach and am currently 48 hours in to the dark frame capture cycle, Astro Pixel Processor can also do dark frame scaling and I have bias master frames at 0 and 100 as that is the two gain settings I will probably use. As this is a new camera and colour the intention is to use the brain to calculate the exposure length but set the gain to 0 or 100 the issue is that the brain will spit out exposure lengths that I would not have so the intention would be to pick the nearest dark that I have and to then scale this during integration.

Having now captured quite a few frames and created the master dark's the main difference I see if doing a full ddp stretch is that the hot or cold pixels increase in intensity as the duration increases the actual background level would appear to be staying very constant so will be interesting when I get to image.

Assuming we ever see the sky again

I also looked at the histogram mean figures, the image below is the value for the upper left quarter of the sensor in layman's terms can you explain what they are telling me this is with the sensor at RAW16 4min exposure and gain of 100 at gain zero they are just under half these values in the right most column.
histogram.jpg
histogram.jpg (7.7 KiB) Viewed 1750 times
Cheers
Nick
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: ASI2600 and Dark Frames

#4

Post by admin »

Hi Nick,

The readouts from the top right of the histogram are the means and standard deviations of the values in each of the colour channels (and in the synthesised luminance signal created by mixing the colour channels).

If you saturate the image by removing the dust cover from the sensor then you should see that the maximum values for the means of the channels in a RAW16 mode is somewhere close to 65,500. The low values that you are seeing indicate a low level of thermal noise in your dark frames. If you varied the exposure of the dark frames then you probably will find that the mean values change roughly linearly with the exposure length.


Cheers, Robin
Post Reply