Your first pipeline
A guided walk from an empty canvas to a filtered, epoched recording. Each step is checked off as you go, and the progress is kept if you leave and come back.
Open a recording
File then Import, and pick an EDF or BrainVision file. If you have none at hand, the Datasets panel can fetch a public one from OpenNeuro.
A loader card appears on the canvas with the channel count and the sampling rate in its title.
Look at the raw signal
Click the output port of the loader card. The visualizer opens on that point in the graph. Scroll through a minute or two: you are looking for drift, flat channels and obvious line noise.
Add a high-pass filter
Drag High Pass from the palette and wire the loader into it. Set
l_freq to 1 Hz. Open the gears view: the transition width
shows as resolved, for example "auto to 413 taps".
Remove line noise
Add Notch after the high pass. Set freq to 50 or 60
depending on where you are, and leave three harmonics.
Re-reference
Add Common Average Reference, or Bipolar Reference for intracranial data. Compare the visualizer before and after: shared noise should drop.
Epoch around events
Add Epoch. Set tmin to -0.5 and tmax to
1.0. If the events table is empty, check the labels on the loader card.
Reject bad trials
Add Reject by Variance with the threshold at 4 standard deviations. Its report output tells you how many trials went and why.
Save the pipeline
Ctrl/Cmd + S. The file is the analysis: archive it next to your data, and anyone opening it runs exactly what you ran.