Long-term recordings, or intermittent recordings?

[elenahindinger]

Hi,

Just looking for some input from more experienced users.

I'm just getting started with stytra and was wondering if anyone has done longer tracking (e.g. overnight, 24h) of freely-swimming fish + visual stimulation, and if so, any words of advice? I'm worried with such long experiments that I won't know whether the tracking has failed or if the fish just stays still for extended periods of time.

Also, is there any functionality to write a protocol to record say for 5min every hour on the hour, things like that?

Thanks!

[vigji]

Hi @elenahindinger! Sorry for the delayed response.

Unfortunately that is not an experimental situation we have ever used stytra in!
I have done some long (12-18h experiments) only with embedded fish. I agree that I would not trusting the freely moving algorithm for such a long stretch without saving the video, but I'm almost sure that experimental length would make video saving impossible.

So, if stytra is particularly convenient for your visual stimulation, what I would advise in your situation is the following: use stytra only to perform the visual stimulation. Use a separate procedure (for example, a custom py script that opens camera, stream frames to a video, closes camera, at regular intervals). Make sure that somehow in the camera video you can "see" the visual stimulus (or a proxy of it, if full stimulus would interfere with tracking) so that it is reasonable to align everything later (it is not very elegant, but could work :) )