LSL stream looks noisy in LabRecorder even though BrainVision Recorder shows clean EEG

[clinicalpsychology180-gif]

I am using a Brain Products actiCHamp Plus EEG system with PsychoPy, LSL, and LabRecorder. The EEG signal looks clean and stable in BrainVision Recorder, but when I view the same data through the LSL Viewer or record it with LabRecorder, some channels appear noisy or show artifacts.

The LSL connector seems to be working properly, and the electrodes also look fine in BrainVision Recorder. The problem seems to happen only in the LSL/LabRecorder view.

What I expected
I expected the EEG view in LSL/LabRecorder to look similar to BrainVision Recorder, with clean and stable waves.

What actually happens
EEG looks clean in BrainVision Recorder.

In LSL Viewer/LabRecorder, some channels look noisy or unstable.

In some cases, only a few channels show artifacts.

Trigger/marker streaming appears to work, but the EEG view still looks different from BrainVision Recorder.

My setup
EEG amplifier: Brain Products actiCHamp Plus

Software: BrainVision Recorder, PsychoPy, LSL Viewer, LabRecorder

Triggering: LSL markers from PsychoPy

Network: LAN, and the computers can ping each other successfully

Things I already checked
Electrodes look fine in BrainVision Recorder.

The LSL connector is running.

LAN connection is working.

Ping between machines is successful.

The issue seems to be only in the LSL stream/view.

Questions
Is it normal for BrainVision Recorder and LSL/LabRecorder to look different?

Should I enable RDA in BrainVision Recorder, or should I use the actiCHamp Plus LSL connector directly?

Could this be caused by channel order, sampling rate, buffering, or timestamp issues?

What is the recommended pipeline for actiCHamp Plus with PsychoPy and LSL?

[cboulay]

Should I enable RDA in BrainVision Recorder, or should I use the actiCHamp Plus LSL connector directly?

That's something to ask your BrainProducts rep. I'd guess the latter?

As for why they look different, I'm guessing that BrainVision Recorder is doing some filtering before visualization whereas when you're viewing the data that LabRecorder saved it is raw and you aren't doing any filtering.

If you can be more precise in your description of the problem then I can be helpful. For example, there is no looking at the data in LabRecorder, so I have no idea what you mean when you say you looked at the data in LabRecorder.

[clinicalpsychology180-gif]

I am using a Brain Products actiCHamp Plus EEG system with a 5-channel montage.

The EEG signals appear normal in BrainVision Recorder, and electrode impedance is good. However, when streaming the data through the BrainVision LSL Connector into BrainVision LSL Viewer / LabRecorder, channels 24 and 32 consistently show large amplitude spikes/artifact-like activity throughout the recording, while the other channels remain stable.

Pipeline:
actiCHamp Plus → LSL Connector → BrainVision LSL Viewer → LabRecorder

Additional details:

RDA was enabled in BrainVision Recorder
The issue occurs continuously on channels 24 and 32
In BrainVision Recorder, these channels look normal
Electrode impedance is good
A notch filter was applied after recording, but the spikes remain visible
The problem appears only in the LSL-streamed data

[cboulay]

I'm sorry but I can't imagine how LSL could be responsible for this. LSL doesn't touch the data. Sometimes LSL will modify timestamps (if you ask it to) so that could lead to weird backwards or forwards time jumps, but it wouldn't create amplitude changes.

Whenever I see behaviour like this it's usually due to some weird wrapping and/or byte-alignment, e.g. due to pushing data from a buffer that includes padding, but I don't see how that would affect only channels 24 and 32. Anyway, if that were the case, that would be a problem on the LSL connector end which is a product of BrainProducts and is not something that can be fixed in LSL.