Exposure time is changing constantly between 2 levels

[ova103]

Hello
evertyhing works fine (PI 4, HQ camera) except exposure time. Let me explain what happened yesterday and today (grey day,)

With these settings (mean target 0.5 and mean treshold 0.05 the exposure remains constant at 939 us.

But when I set the mean target at 0.6 and mean treshold still at 0.05 the exposure is switching between 558us and 1463ms.
See screenshot and log file in attachment and 2 consecutive pictures

allsky.log

Why is the exposure switching like that? I'd like to understand so I can finetune my settings better.

Thanks for the feedback!

PS: I changed mean target back to 0.5 and now exposure time is switching too.

[EricClaeys]

Are you saying it's now working fine?

[ova103]

No, it's switching exposure all the time

[EricClaeys]

Please follow the instructions on the Troubleshooting -> Report Issues page and then attach /var/log/allsky.log.
Screenshots of the file aren't very helpful since I can't search and scroll.

This is the first time I've seen the exposure "ping pong" on an RPi camera (too bright, then too dim, then too bright ...).
The RPi auto-exposure algorithm has been rock solid.

[ova103]

Hello on my first mail I have added this file. Do you see it? If not I'll send it again.

[EricClaeys]

The instructions on the documentation page have you truncating the log file and then attaching the new one after the problem happens.

Most of the entries in the log file you attached weren't with a high enough Debug Level.

[EricClaeys]

As a side note, I suggest setting your daytime Manual Exposure to 0.5 which is similar to a typical daytime exposure. Using 50 causes Allsky to take quite a few frames trying to find the correct exposure.

[ova103]

Ok I'll try this out tomorow

[hoetzgit]

+1 with this "ping pong" problem during daytime

[EricClaeys]

@hoetzgit,
Please follow the instructions on the Troubleshooting -> Report Issues page and then attach /var/log/allsky.log.

Are you also running v2024.12.06 ?

[ova103]

Hello

First best wishes for 2025 and thanks a lot for the Allsky developments and support.

Yes I'm running on version v2024.12.06

I did not yet change to daytime exposure to 0.5. I left it to 50 to see if the problem appears again. And yes it is.
The exposure time is not constant. But this time it is not doing ping pong between 2 values, but to more than 2 values.

Here are the files:
allsky.log
settings.json.txt

Hope this is the right procedure I followed.

Thanks for the support.