Vertical Banding ToupTek ATR3 CMOS 26000 KPA · Eric Maurício · ... · 9 · 368 · 6

Hello all, I hope you're doing great.

I've purchased the Touptek2600c (tsoptics rebrand) around 2 or 3 months ago, as per usual, the weather wasn't cooperating, so I had to wait arund a month before I could start shooting. My plan for clear nights with little to no Moon, was to shoot the Shark Nebula, and so I did. 

After I got around 6h of data, I started to have a go at it in Pixinsight, and the problem first appeared. As I apply the Auto Stretch, I noticed vertical banding with alternating magenta and cyan cast, like so:


This is a 18h stack, with no calibration frames, with ABE and Spectrophotometric color calibration applied, StarXterminator then saturated to enhance the issue. 
My acquisition process is as follows: 300s subs, 100 gain, 50 offset, at -15ºC. High Gain Mode enabled, Dew Heather on, although I don't think the camera as one, Ultra Low Noise Enabled. I dithered every frame.

Solutions that I've tried: Obviously I started by including calibration frames, darks, flats, and dark flats, none of them resolved the issue. Then I thought that, since I don't get the banding using the L-Ultimate filter, it could be a filter problem, so I removed the Svbony IR/UV cut filter, tossed all the frames that I had previously taken and shot again. Ended up with the picture above, so I guess it isn't the filter, but I cannot reproduce the issue, using the exact same acquisition process and stacking, with the L-Ultimate filter. 


This is around 9h with the same acquisition and stacking technique. 

Since I'm aiming at an LDN, I thought that it could also be that I'm clipping my blacks, so I increased the offset to 200, same result.
Then, I tried with 60s, and after that, 120s subframes, because I could probably be saturating the pixels and getting the pattern as a result, nothing changed.

I debrayed a single frame, stretched it, and the pattern (I think), is barely visible.


My theory is that pixels that don't get enough signal tend to form a patter noise with those colors, it's the only thing that I can think about, but I have no idea anymore, dark frames should enhance the issue and correct it.


300s dark frame, debrayed, stretched.

If anyone as any idea what the problem seems to be, or as any input on the situation, I'd love to hear it.

Thank you and Clear Skies.

Hi,
if you use NINA, you can try to decrease the USB-Speed to 2 or 1.

Best Regards
wittinobi

USB speed was set to 2, will try 1, hopefully I can get some hours tonight and see if it helps.

Thank you!

Hi Eric! Did you manage to resolve the issue? I'm having the same problem with the same camera

Maxim:
My banding pattern is visible the best on flat images(mind the dust speck, I've just unscrewed and screwed the camera). The worst part that it's not so strong(if exists at all) on the light images, that's why when I stack without flats, the image is better off, at least no visible banding is there. But these kind of flats ruin the stack, while correcting the vignetting and the dust. I've tried all the possible settings, usb speed, readout mode, ultramode etc. Even connected it to usb 2.0. No success whatsoever.

What is your exposure time for the flats? Too short of an exposure time can, on some cameras from some manufacturers, cause the camera to enter a different mode. QHY's drivers call this the "Stream" mode, for example. A common rule of thumb that I've seen is that you should aim to have your flat & bias exposure times above 0.2s to avoid these modes, though the exact cause (and exact minimum exposure duration) are not fully known.

(For the nerds: Note that this is distinct from the camera's readout mode. Stream mode is a separate phenomenon. This is also different from the ASI1600's weird behavior for exposure times under 0.2s).

Maxim:
Hi Eric! Did you manage to resolve the issue? I'm having the same problem with the same camera

Hi, Maxim!

I worked out the issue by correcting my flat frames, which I didn't at the time, or didn't do correctly, the banding appeared only in the flats, which, if left uncorrected, would result in the banding. So I took darkflats after my flats, stacked them as "bias frames" in Pixinsight, and managed to eliminate the banding.