Event-Triggered Analysis¶
Plugins for the Open Ephys GUI that analyse continuous data in windows locked to an event — a TTL edge or a broadcast message.
| Plugin | Appears in the GUI as | What it does |
|---|---|---|
| Triggered Average | Triggered Avg |
Time-domain average and standard deviation, with individual trials |
| Receptive Field Bar Mapper | RF Barmapper |
Visual receptive fields, back-projected from the per-direction trial averages of a sweeping bar |
| Triggered Power | Triggered Power |
Power spectra accumulated across trials and split by condition |
| Triggered Coherence | Triggered Coherence |
Work-in-progress: Magnitude-squared coherence and coherency phase for configured channel pairs |
Triggered Coherence is work in progress
It builds, loads and computes, and the estimator is tested numerically — but it has no pre-trigger baseline (#16) and its pair edits are not undoable (#14).

What they have in common¶
All four sit on the same capture layer:
- The same trigger model. A trigger source is one experimental condition. It names a TTL line and, optionally, three broadcast-message patterns — arm, cancel and commit. See Triggers and messages.
- The same trial window.
PreandPostmilliseconds around the trigger, plus a channel selection. Cost is linear in the number of selected channels. - The same three editor buttons.
TRIGGERS(the condition table),MONITOR(per-source counters) andANALYSIS(everything that reshapes the accumulators). - The same trigger-table file.
SAVEandLOADin the TRIGGERS popup move a whole condition table between plugins. - The same session format. Triggered Average and the Bar Mapper write a directory of
one XML manifest plus one
.npyper array, readable in Python and MATLAB.
Where to start¶
-
New here — Installation, then Getting started.
-
Setting up conditions — Triggers and messages.
-
Looking up a number — the Parameter reference.
-
Analysing the output — Saved sessions, and the Python and MATLAB loading guides.
Notes¶
- All per-trial work runs in the background, off the data-acquisition path.
- There is no decimation here. Put a downsampling plugin upstream if you want to analyse a reduced sample rate.
- Coherence is only meaningful pooled over trials — a single trial has coherence 1 by construction.
- The receptive-field mapping is recomputed in the background, so map settings can be changed without recapturing anything.
Licence¶
GPL-3.0-or-later. Copyright © 2025–2026 Joscha Schmiedt, Universität Bremen; parts © 2022 Open Ephys.