Exposure times in a sequence rounding?

Hi -

Still learning SGP - so far, I’m a big fan.

SGP v3.0.3.140. Using ASI1600MM-PRO camera. I can provide plenty more details if they are pertinent…

When I set up a sequence, I tend to use a single target to cover LRGB, then also have Flats, Darks, and Dark Flats all as separate events. Due to details not much worth discussing, the correct exposure time for my flats are things like 0.006, 0.012, 0.015 seconds. I find that when I save these into a sequence, sometimes (but not always) they get rounded after I stop / restart (or maybe some other event) the software. 0.006 and 0.012 become 0.01, for example.

I suspect I’m far from the first person to walk this path… Am I doing something wrong?


  • Matt

I don’t think SGP has thousanths of a second, so it would round up. The bigger question I have is why are your flats so super short?

A) The dimmable drawing table I bought off amazon dims using PWM, which is visible as striping when I take flats. As a result, I have to take them at the drawing table’s full brightness (no PWM engaged), with 4 layers of white T-shirt. I could probably T-shirt more if necessary, but…
B) SGP is sensitive to thousandths of a second - I see a reliable, reproducible result when I change a single thousandth in the scenario I described. SGP even seems to remember the thousandths setting - sometimes. Hence this post. :slightly_smiling_face: At the values I mentioned I land in the ~17-18k median value range, which is (I’ve read) a good target for flats (25-30% of max value).

  • Matt

I see; so I was wrong about the thousandths of a second. So if there is indeed a problem with SGP not remembering the exposure setting, can you post a log file where this happens?

What happens if you push the ADU to 25,000? Can you get the exposure to .01? I don’t think it will make much difference using flats that are 25,000 instead of 18,000 (I have a QHY163M and have used anywhere from 18-28,000).

I don’t know SGP well enough to know where / how to find said log files. I’d be happy to try and reproduce. Can you give me guidance there?

  • Matt

Oh, and to answer your question - 0.01 sec with my L filter at unity gain (with 4 layers of T-shirt) reads about 36k. I’m sure I could layer more T-shirt, but this approach seemed to work.

I’m not very good at reading those log files, but hopefully someone else will chime in.

Yeah, 36k is too high. You must have more patience than I because I would have either returned or thrown away that drawing tablet a long time ago. :rofl:

Just an update -


I saw the values change directly. It was in response to disconnecting the camera at the end of the night - when I picked the menu option to disconnect all equipment was when I saw my 4 flats targets duration change - they milliseconds were dropped, and the values were rounded to hundredths of seconds.

I suspect the event occurred circa 00:54:27.597 on 12/8/18, line 78916 of the file named sg_logfile_20181207204022.txt.

Just dusted off Dropbox for the first time in many years - please let me know this worked!

  • Matt

Can someone from SGP please confirm that I’ve provided logs in a way that is effective?

  • Matt

One more ping here, hoping for a response…

Sorry this fell through the cracks (more than once :frowning:)

I’ve found and addressed the issue. SGP will now allow ten thousands place for all exposures in the sequencer.