search

neurogears / vestibular-vr

Closed-loop VR setup for Rancz Lab
2 stars 0 forks source link

Main workflow - VR not running when runGain and flowXtovisual gain set correctly #71

Closed EleonoraAmbrad closed 2 months ago

EleonoraAmbrad commented 3 months ago

recently, we have experienced strange behaviors from the Main workflow. When using a session settings file that we know works to produce the desired VR motion etc, sometimes (apparently randomly) the VR does not move even if the flowYtovisual gain is >0 and the ball is moving and flow sensor is producing values. This only happened at the start of a block (i.e. it does not occur in the middle of a block), but it has occurred after 2 blocks had already ran successfully. Whether this happens or not doesn't seem to be dependent on block length.

We are running Main experiment workflow from main branch.

On the bonsai console there does not seem to be any error when this happens.

I am wondering if it is a similar issue as the one that was causing H2 to crash (see issue https://github.com/neurogears/vestibular-vr/issues/63).

Here is a csv file with the flow sensor data from an experiment where the VR did not start for the third block, which occurred at around 20 min in the session. It doesn't look like anything strange happened to the flow sensor data, except some strange clipping at 0? playback_file_B3M3_MMreg.csv

H1 does not seem to be having timestamp issues either, but let me know if you need the original data.

Thanks for the help!

RoboDoig commented 3 months ago

There are 2 subjects, RightPhaseProcession and LeftPhaseProcession that should be within the CalculatePhaseProcession sub-workflow. Can you check if they produce any values when the Vr motion stops?