Order of the raw EEG data

User avatar
James
Site Admin
Posts: 1103
Joined: Wed Jan 02, 2013 9:06 pm

Re: Order of the raw EEG data

Post by James »

Bundles are the standard, so I'm not going to make a note about that sorry. I know that the common python libs don't support bundle timetag, but all other languages do.
The bundle timetag is not super accurate as it's generated on the phone, so it's prone to Bluetooth packet buffering. You will see the time Delta is not consistent.
I have complained about this to Interaxon, but they so far have not updated the firmware to output the time. There are quite a few posts about this on the forum.
stellarpower
Posts: 10
Joined: Sat Jul 08, 2023 12:36 pm

Re: Order of the raw EEG data

Post by stellarpower »

Yes, I had been using this repo, which uses python-osc, and I think that must abstract the bundle aspect away. Hence why it took a bit of fiddling to work out the format, as I moved it into Crystal and the messages I'd dealt with before using the same library there were all just single messages, without using the bundle feature. Even using pypy it was taking up massive amounts of CPU and currently working on my laptop I noticed a few days ago there was a lag of nearly two seconds, you'd have thought they would wrap a C library or similar but I believe that pip package is implemented in pure Python. Might be worth recommending people try something else if they're working with their Muse data.

Also a shame the firmware doesn't output the timestamps, as having that at source would be the simplest way to know you have the best-quality data. Then any latency is less important as you can in theory reconstruct the data on the other side. Oh well. Thanks for clarifying on this!
Post Reply