You have bad sensor contact. It's worst on the ear sensors, but also bad on the forehead sensors. The flatlines are when the data becomes unusable.
To check your data quality, switch to the raw eeg view in Muse Monitor. On this graph you should see thin crisp lines, with well defined spikes when you blink. Just like in this screenshot from the home page: https://musemonitor.com/img/mm-screen-3.jpg
To get good sensor contact you must have an electrically conductive path from your skin to the sensor; this means a thin band of salty sweat must form. If you skin is dry, or hair is in the way, then you will get a bad signal. To help it get started I recommend that you dip a finger in clean water and wipe the sensor band to get it moist.
Thank you so much for your comments about bad sensor contact.
I truly understand what you are saying but we have used that "clean water" technique many times with no success.
It still drops out at EXACTLY the same point for both of us, circa 130. Every time, over dozens of tests, all with the care to get good sensor contact exactly as you say.
We approached a neuroscientist who thinks it could be to do with the MUSE sampling rate of 256 HZ as follows.
What do you think? This is from him.
"It's true that flat lines usually indicate signal loss, but my suspicion is that in your case, this is not a signal loss issue. I think the core of this issue could be due to what's called the sampling rate of an EEG. Typically, you need double the sampling rate of the highest brain frequency you'd like to study to effectively get data on the frequency. For example, most full EEGs have a sampling rate of either 250hz or 500hz. A 250hz sampling rate basically means the electrode is taking data 250 times/second and is capable of reliably measuring up to 125hz of brain activity. Anything more and it is too much to capture. I don't know the sampling rate of the muse or the exact frequency of your gamma activity, but it is completely possible that your brain is simply too fast for the muse to read, which makes sense if you look at the graph. It rises steeply, gets that same relative marker of around 130 and then all of a sudden flat-lines (your gamma is likely going "off the charts" at this point) and then once you lower the energy back down it begins to read it again once it reaches 130 and goes lower. I'm not certain if this is the truth quite yet, I'd look into the csv file and check for signal loss."
