Vertical Banding - What am I doing wrong?

[Jammy]

I've been trying to image M51 with Evostar 80ED and ZWO ASI1600mm Pro cooled camera.

I've used the camera at unity gain at -10C many times and it's always worked really well.

As an experiment, I've decided to change the gain to zero to see what difference it makes.  I've also added a IDAS D2 LPS filter and increased my exposure time from 20 secs to 60 secs.  In bortle 6/7 skies data was blowing out above 20 secs of Lum before I got the LPS.

All my images are now coming out with vertical banding.

Is this due to the camera gain value?

I've added some shots of what I'm getting out of the camera:

Single Light Frame

Master Dark

Master Dark Flat

Master Flat (ignore all the dust spots!)

Output file from APP without any further editing.

My flats and dark flats are taken with a light box at around 1/10th sec, is that too short?  I'm getting the required 20000 ADU value at that for Lum.

I've not noticed any problem with the camera at unity gain,  but having seen other people using gain zero for images and thought I'd try it.

Any help would greatly be appreciated, thanks.

[Knight of Clear Skies]

I'm don't understand this well but I think you've effectively boosted up the fixed-pattern noise from the sensor by running at zero gain, boosting gain reduces read noise. I don't believe there is any reason to run below unity gain unless your images are over-exposing. (Below unity gain I think your camera is discarding real data.)

Edited April 17, 2020 by Knight of Clear Skies

[Jammy]

3 hours ago, Knight of Clear Skies said:

Below unity gain I think your camera is discarding real data

I didn't know that.

I've only tried because I've seen other people getting good results at gain zero.

I'll be completely honest, I've no idea what it is supposed to do if anything.

I've asked the same question on one of the ZWO forums and I've been told the camera does not perform well at gain zero and it's been suggested using at 75 gain.

I tried it at 75 gain last night, but I'm just running off some darks and flats to see if its any better.

I did the Sharpcap sensor analysis and smart histogram and it gave 138 for unity gain (ZWO say 139), and 59 for high dynamic range value.  I'm not sure what HDR is or does.  I'm a novice when it comes to these type of cameras.

[jimjam11]

The banding is most obvious at gain 0 but it calibrates out fairly easily. 
I always use gain 0 for L, and gain 76 for RGB. That gives me 60s subs for both which is optimal in my sky (bortle 5/4). If I used gain 139 I would be down to <30s for rgb and <10s for l.

If you use PI create a master bias and then run the super bias process to generate an excellent master bias frame. If after calibration there is any banding left use the canon banding removal script.

 

Edited April 17, 2020 by jimjam11

[Jammy]

Thanks for the response.  I've not used PI, and can't really comment.  APP doesn't cope very well with the banding and that's why I've noticed it.

I have the same issue even at bortle 6/7.  My L subs start to blow out above 20 secs, that's why I wanted to lower the gain value.

[jimjam11]

If you are bortle 6/7 then your subs could be impractically short. This is a brilliant reference for the asi1600 in terms of exposure:

https://www.cloudynights.com/topic/573886-sub-exposure-tables-for-asi-1600-and-maybe-qhy163/

 

Also watch out for your imaging efficiency with the asi1600, I have really had to work hard to avoid wasting tons of time. See this thread I started a few days back on the exact subject:

 

With your skies it seems like a no brainer to use gain 0 for broadband imaging unless you want to stack potentially thousands of subs...