[FieldTrip] saving events to file in realtime processing

[Alejo Salles]

hi,

I'm looking into doing EEG realtime processing using biosemi active two
hardware. I use the biosemi2ft program to acquire data and create the
buffer, from which I read from matlab. as the processing I need to do is not
very intensive, while it's important to keep a low time lag, I'm running
matlab on the same computer that does the acquisition.

I would like to know how to go about sending events to be recorded in the
gdf file. firstly, I'd like to know whether this is the right way to go, or
if I should instead have the buffer spawned by the matlab code... can I
write to the buffer from matlab even if it was spawned by biosemi2ft?

supposing this is ok, how should I send the events? I understand that I
should set event.sample, but I'm not sure what to put here, should I use the
one after the one just read? how can I make sure this will get recorded to
file? should I use a fixed delay instead?

if I spawn the buffer from matlab instead, how can I then tell biosemi2ft to
write the events to file?

is any of these the way to do this or I should resort to send signals
through the parallel port into the biosemi box?

on a different note, is there a way to downsample the rate for saving the
file in biosemi2ft? I gather the setting is only for streaming, but then I'm
limited by the biosemi hardware to sample rates >= 2kHz, is this right?

many thanks,
alejo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20111027/f72271f9/attachment-0001.html>