8
Beware the real time calibration drift in consumer EEG headsets
I keep seeing people in forums treat the default calibration from Muse or NeuroSky as if it's a fixed setup, but after 6 months of logging my own sessions I noticed baseline readings were shifting by 15% just from humidity changes in my apartment. That means the alpha wave feedback you're relying on for focus training could be completely off unless you re-calibrate every single session. Has anyone else checked their raw data against a known reference like an OpenBCI board to see how bad the drift actually gets?
3 comments
Log in to join the discussion
Log In3 Comments
miller.avery15d ago
Yeah, I saw the same drift. Recalibrating before each session fixed it completely for me.
5
tessa39415d ago
Wait 18%?! That is genuinely insane and I honestly had no idea humidity could mess with these things that much. I always just figured calibration drift was my headset being old or something, not the actual air in my room. Like, I sit in a super humid basement sometimes and my scores are always wonky, now it totally makes sense. That UC San Diego paper sounds wild, I might have to look it up just to wrap my head around the numbers. Thanks for dropping that knowledge, seriously eye opening.
4
jana_hernandez15d ago
Read a white paper from a neurotech lab at UC San Diego that tested this exact thing. They ran Muse headsets alongside an OpenBCI Cyton board in a controlled environment and found the consumer units had calibration drift that got worse with temperature and humidity shifts over 20 minute windows. The alpha band power readings would slide up to 18% off baseline in high humidity rooms. You are spot on about needing to recalibrate every session, but people dont realize their environment is the real culprit. I started doing a quick 30 second baseline reset before each use and my focus training scores finally stopped jumping around randomly.
-1