tl;dr Restarting the MEG electronics turns off the projector “pixel mode”
Tasks that use the projector often use the “pixel mode” to send visually synchronized markers to the MEG. We found that when restarting the MEG, e.g. because of electronics errors, the projector controller (aka Datapixx controller, blue box) resets some settings, turning off the pixel mode, probably because the voltages on those digital channels temporarily change on the MEG side.
As such, markers may be missing if your code doesn’t set it again after restarting the MEG. This could happen for example if your task runs multiple blocks within a loop in your script, but the MEG acquisition is stopped and restarted for each block. If you only configure the projector at the start of the task, the settings could change between blocks.
It is therefore highly recommended to design your code so that projector settings are set at the beginning of each MEG recording, either by running your script separately for each block, or including the setup at the start of each block.
I’ll also bring this to the attention of VPixx.