Deep Sky Sequencer - captured frames bad

[ChrisR Oz]

Hi Robin,

Tried the Deep Sky Sequencer to manage a simple mosaic (SC v4.1.11786). I wanted to capture a fixed number of frames (33) at 30s exposure and 50 at 4s exposure, with the aim of making a HDR composition for a 4 field mosaic of M42, and simply repeating the sequence for more total exposure. The sequencer seemed to work OK, slewing to each field, plate-solving and re-centering and acquiring the frames at 30s and 4s. However, when I checked the frames many at 30s were bad, and most at 4s were bad. By bad I mean with strange banding, like part of the image area was missing its Offset.

I was just using ASI Fits Viewer to look at the FITs files. The first 4s file already has problems (see link: https://drive.google.com/file/d/17wQ07x ... sp=sharing)

ASI FITs View of first 4s frame from Sequencer

Sequencer problem1.jpg (156.5 KiB) Viewed 287 times

But I can see it also in SharpCap, using Folder Camera mode and Live Stack. This example is for 4s exposure and I have selected a 4s Dark and a suitable Flat frame.

First frame loaded as Folder Camera in SC Live Stack

Sequencer problem2.jpg (491.61 KiB) Viewed 287 times

Returning to a single frame captured using the folder camera, I can use the Smart Histogram to probe parts of the image. The bottom part shows a peak, even away from stars, as expected for Offset/Backgnd = 5 (this is a ZWO 294MC Pro). However, the top part is missing the offset, and is black between stars, with no general offset.

This is my first foray into Still Mode. I usually use Live Stack, as it's always a great experience seeing a galaxy emerge in real-time. Perhaps I am missing something about using Still Mode.

Are these bad captures an indication of a SC bug, something wrong with ZWO API interaction or something I have setup incorrectly?
Cheers, Chris.

[admin]

Hi Chris,

I think the first thing to check is whether you had any of the 'preprocessing' controls in SharpCap active while you were capturing these images (subtract dark/background subtraction/flat correction/etc). Those will be recorded in the 'CaptureSettings.txt' file saved next to the captured images. Those controls are the only things that should affect the image data between it coming off the camera and getting saved into the output file when doing simple capturing, so if those are all 'Off' then most likely the images came off the camera (or at least out of the ZWO SDK) in that state.

If some of the preprocessing settings were on then we would need to investigate those, but to get images like that would probably require weirdly banded flats or darks during capture, which would seem unlikely.

I'm afraid that I suspect the most likely thing here is a hardware issue with the camera or a USB/power cabling issue. Looking at the FITS file that you shared there is a 'normal' section in the middle, a 'dark but some stars' section at the bottom with practically no background but some bright stars and a thin 'completely black' bar at the top where all the pixel values are zero.

I would be tempted to try some test exposures under very dim (but not zero) illumination to see if the problem persists and if camera setting changes have any effect on it (cooler on/off, Turbo USB setting, etc).

I think that still mode/live mode is unlikely to be the issue here (or the sequencer for that matter). SharpCap runs ZWO cameras in still mode behind the scenes when the exposure is 2s or above anyway (even in live mode), so if you set a 4s exposure on a ZWO camera but leave SharpCap in live mode, SharpCap basically requests 4s still frames from the camera one after another...

cheers,

Robin

[ChrisR Oz]

Hi Robin,

As I noticed this problem during acquisition, I immediately followed the Still Mode sequence with a Live Stack, with frames saved. All Live Stack (30s and 4s) frames are perfect, as usual. Indeed, the order of events was Still 30s and 4s, Live 4s, Live 30s. All live frames are fine.

Yes, I did have pre-processing set. I did not realize that the saved Still frames would have these applied, as Live frames have nothing applied (just the stack as viewed has these applied). However, that does not really explain the banding pattern.

You say that 4s and 30s exposures are "effectively" done in Still mode. However, the results are very different ...

I guess I'll stick with Live captures.
Cheers, Chris.

[admin]

Hi,

there is an odd quirk in the way saved frames include/exclude processing in SharpCap - at least it can seem a quirk at first sight.

* If you are capturing normally using 'Start Capture' etc in live or still mode then the saved frames *include* the effects of any preprocessing settings (dark, flat, background subtraction,etc)

* However, if you are *Live Stacking* (Deep sky version to be precise) and saving raw frames then those saved raw frames *do not* include the effects of preprocessing settings.

I don't know if this could be part of the explanation of why you see different behaviour when going back to live stacking. It would still require weird flat/dark frames or some other glitch in the processing though, which would need to be explained properly.

As for running the camera in still mode for longer exposures - it turns out to be easier to deal with things like stopping an exposure in progress and starting another in response to things like changing the exposure value if the camera is in still mode for long exposures - the capture is also more robust too. The upshot of that is though that the way SharpCap talks to your ZWO camera is the same regardless of SharpCap's live/still mode providing the exposure is above 2s, so the actual switch of SharpCap between live and still mode is unlikely to be the underlying cause.

cheers,

Robin

[ChrisR Oz]

I guess I’ll just give up on Still mode, as Live stacking works fine.
Cheers, Chris.