getting time series through OSC in PureData
in OpenBCI_GUI
Hi,
I want to use OpenBCI datas in PureData. No problem to get them (although I had to guess and retro-engineer to get the labels, not really documented).
I get series of 10 values per channel in each package. I assume they represent 10 instant value, so:
-Sample rate is known to be 250Hz, confirmed by the time calculation: 1 package of 10 values / 40ms ( ( 1 / 0.04 ) * 10 = 250 ).
Well now I guess I have to sequentialize the lists of values every 4ms, right?
-
Comments
Mentioning Richard @retiutut, who has used Max / PD in the past.
Are you referring to the GUI Networking Guide link on this page, (scroll the document to the UDP section),
https://docs.openbci.com/Software/OpenBCISoftware/GUIWidgets/#networking
Multiple 'samples' are packed to improve network bandwidth utilization. You control what data is sent.
Cyton ADC samples every 4ms (250 Hz) and collects 8 EEG channels and 3 Aux channels (accelerometer or other info) on each sample. The GUI Networking Widget lets you select what preprocessing occurs before OSC packets are sent.
William
Happy to review this! Can you please share the full PD file for this? We have a related open ticket in our Docs, and resolution in this forum thread is a big step towards documenting this. https://github.com/OpenBCI/Documentation/issues/169
It's not fair, the document changed since last week
it's clearer now.
Thank you for your answers. I finish the sequence part to output a dataflow and I'll post the patch for sure.
Here is a first version. Any comment welcome.
(actually I can't attach PD file: I chose the file but nothing happens...)
ok, it' seems I'm not allowed to post files here, so:
http://rph-r.cc/rph/OpenBCI.pd
by the way: what are the max/min values? It goes higher and higher (>45K). The resolution is impressive.
The raw time series includes a DC offset, which is in the millivolt range. Tens of millivolts is tens of thousands of microvolts. Normal EEG is between a few uV to say 80 to 100 uV.
You must remove the offset for most EEG applications. This is done with a highpass at around .5 Hz, or a bandpass from .5 Hz to say 45 Hz. Frequently a notch filter at your mains frequency is also advised.
https://openbci.com/forum/index.php?p=/discussion/201/large-millivolt-data-values-fbeeg-full-band-eeg