Hello,
We are attempting to perform a closed-loop experiment, where we send continuous buffers of data to Python using the ZMQ interface plugin, process them to detect an electrophysiological pattern, and then use the NetworkEvents plugin to trigger TTL events in the GUI. EventsBroadcaster is there as a safeline that sends event timings back to Python to be saved separately. (signal chain attached).
We are observing a really puzzling problem:
The TTL events seem to show up as delayed relative to the stimulation, which impedes adequate characterization (as shown in picture - the spikes correspond to stimulation artefacts, and should be after the TTL...):
I checked the sample number of the continuous data (it begins in 0), and also looked at several ways of reading the data (both through the provided analysis tools and also other methods). There seems to be an around 250 ms delay to every TTL...
I can provide more information and the data by e-mail, if helpful to solve the problem :D.
Hope you can provide some valuable insights,
Best Regards,
Pedro