Fixing tests related to image solving with simulator

[wtgee]

• Use unsolved image so we can test solving
• Pass skip_solved=False during analyze_recent. Helps with testing when
  the same image is used repeatedly, shouldn't affect normal usage.
• Add some extra logging
• Misc cleanup

[codecov]

Codecov Report

Merging #627 into develop will decrease coverage by 0.02%.
The diff coverage is 58.33%.

@@             Coverage Diff             @@
##           develop     #627      +/-   ##
===========================================
- Coverage    67.22%   67.19%   -0.03%     
===========================================
  Files           65       65              
  Lines         5690     5692       +2     
  Branches       798      798              
===========================================
  Hits          3825     3825              
- Misses        1655     1656       +1     
- Partials       210      211       +1

Impacted Files	Coverage Δ
pocs/camera/simulator.py	100% <100%> (ø)	⬆️
pocs/scheduler/dispatch.py	98.03% <100%> (ø)	⬆️
pocs/observatory.py	76.1% <28.57%> (-0.46%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2f6f5fb...aa15829. Read the comment docs.

[AnthonyHorton]

Code changes look fine to me, so if it's solving the problem it was intended to solve LGTM.

[AnthonyHorton]

An aside on the image solving, we noticed during the recent testing with Huntsman that if given ludicrously large images the solving will timeout. We saw this consistently when one of the 50 Mpixel FLI ML50100 cameras was enabled, which due to its name would become the default primary imager. We got a warning in the log about analysis failing due to solving timing out for every image we took.

[wtgee]

An aside on the image solving, we noticed during the recent testing with Huntsman that if given ludicrously large images the solving will timeout. We saw this consistently when one of the 50 Mpixel FLI ML50100 cameras was enabled, which due to its name would become the default primary imager. We got a warning in the log about analysis failing due to solving timing out for every image we took.

There is a timeout param than can be passed through get_solve_field that will let it keep going. I use that on the Pi where the solve can take 45-60 seconds. Default is 30.

But we should probably figure out if the solving can be done asynchronously (and/or if needed at all for Huntsman observing).

[AnthonyHorton]

There is a timeout param than can be passed through get_solve_field that will let it keep going. I use that on the Pi where the solve can take 45-60 seconds. Default is 30.

I suspected that there was a parameter for this somewhere.

But we should probably figure out if the solving can be done asynchronously (and/or if needed at all for Huntsman observing).

Definitely. The processing that goes on at the end of each science exposure can take quite a while. It's not a huge issue for 5+ minute long galaxy exposures but it is contributing to some substantial gaps in temporal coverage when trying bright star transits. Should look at deferring as much as possible to after the next exposure has started.