Time to take image differs from requested exposure time.

[kalalaa]

Dear author, I have a problem with my allsky camera. I use a T7C with ASI compatible firmware. It worked for years, but my raspberrys sd card failed, so i reinsalled the whole thing. Now after dark subtraction, the final image is really noisy. I viewed the log, and the exposure time differs from requested. I tried manual USB bandwidth with different values, tried 0.8 exposure off, tried different exposure times in manual, and everything.

Sometimes the darks are different from eachother. So a 10°C and a 11°C has a really different noise pattern.

From log:

20 sec night:

STARTING EXPOSURE at: 2022-03-31 21:05:45 @ 20.0 sec Mar 31 21:05:54 allsky allsky.sh[6824]: *** WARNING: time to take image (8652889) differs from requested exposure time (20000000) by -11347111, threshold=10000000 Mar 31 21:05:54 allsky allsky.sh[6824]: > Saving NIGHT image 'image-20220331210545.jpg' Mar 31 21:05:54 allsky allsky.sh[6824]: > Sleeping 30.0 sec from manual exposure

This is three different 20 sec dark:

STARTING EXPOSURE at: 2022-04-01 16:11:52 @ 20.0 sec Apr 1 16:12:02 allsky allsky.sh[3661]: *** WARNING: time to take image (9639824) differs from requested exposure time (20000000) by -10360176, threshold=10000000 Apr 1 16:12:02 allsky allsky.sh[3661]: > Sleeping 30.0 sec from dark frame exposure Apr 1 16:12:02 allsky allsky.sh[3661]: > Saving dark image 'dark.jpg'

STARTING EXPOSURE at: 2022-04-01 18:30:04 @ 20.0 sec Apr 1 18:30:07 allsky allsky.sh[4631]: *** WARNING: time to take image (2234520) differs from requested exposure time (20000000) by -17765480, threshold=10000000 Apr 1 18:30:07 allsky allsky.sh[4631]: > Saving dark image 'dark.jpg' Apr 1 18:30:07 allsky allsky.sh[4631]: > Sleeping 30.0 sec from dark frame exposure

STARTING EXPOSURE at: 2022-04-01 18:34:50 @ 20.0 sec Apr 1 18:34:59 allsky allsky.sh[5162]: *** WARNING: time to take image (8676416) differs from requested exposure time (20000000) by -11323584, threshold=10000000 Apr 1 18:34:59 allsky allsky.sh[5162]: > Saving dark image 'dark.jpg' Apr 1 18:34:59 allsky allsky.sh[5162]: > Sleeping 30.0 sec from dark frame exposure

Why have i different exposure times? Thank You!

[EricClaeys]

@kalalaa You have two unrelated thins going on. When did you take your darks? If it was more than about 9 months ago, you'll want to retake them. The algorithm to take them and to subtract them has improved.

For the second issue, I don't know what causes some ZWO cameras to apparently take exposures shorter than requested. That check is new. If you look at the timestamps in the log you posted, you'll see the software requested 20 second exposures but returned in significantly less than that. When that happens, do the pictures look OK? Try turning on Show Mean Brightness in the WebUI and see if it's jumping all over the place.

[kalalaa]

My darks are new. Maybe the big difference in darks came from frame splitting. Now i have some good darks, with acceptable final noise. When i take new darks with different temperature, i check them. If the new darks have different pattern, i delete them, and capture it again, so i now can handle it. But you can see, sometimes the dark is only 2.2sec and sometimes it is 8.6 sec long. When the subtraction is on, with correct darks, there is a little noise on the final image. I think this come from the different dark, and light timings. The final problem is this. I found some ASI forums, with long exposure issues, so the problem is maybe in the asi firmware. The final image is acceptable, with some noise, and i can use it, so not a big problem, but in older version, i didn't have this.

Thank you!

[kalalaa]

I found in other forums, there are problems with ASI120 long exposures, with some softwares. https://bbs.astronomy-imaging-camera.com/d/6968-zwo-asi120mc-raw16-long-exposure-linux https://www.cloudynights.com/topic/487175-zwo-asi120mc-for-dso-photography/

[kalalaa]

Oh and i found this. I will try it, but will the newest allsky work with this? https://github.com/rkaczorek/astroberry-server/issues/109

[kalalaa]

I installed debian buster, and the allsky. Timing has the same behavior. I found an interesting thing: If i turn off 0.8 exposure, the time has negative difference (at 20 sec the exposure is about 8 sec), and if i use 0.8 exposure, the difference is positive. At 20 sec the exposure is around 38 sec. BUT the final image has around the same brightness, so the real exposure is 20 sec in both cases. The real difference, when i use 0.8 expo, the noise is lower, and stable, when 0.8 off, the noise pattern sometimes different from other darks.

Edit: With 0.8 exposure off, the dark frames are splitted. Thats why is the different noise pattern. Now under sky test: Dark substraction works, but looks like it is overcorrecting a little. Noise almost invisible on dark sky, only some purple and yellow pixels is visible when it is cloudy. I know, the noise is rgb, so i think the this is because the jpg compression. Not a big problem.

You can close this thread! Thank you!

[EricClaeys]

Closing Issue, per user request. Problem solved to his satisfaction.