Which EEG correlates does the app use to determine "calm" vs. "active" states?



How does the app determine “calm” vs. “active” states? Is it looking for e.g. increases in alpha and theta power [Fell et al., 2010] or lower trait frontal gamma but higher trait posterior gamma power [Berkovich-Ohana et al., 2011] or some other published effect of meditation on EEG?

I would also be interested in whether there is a way of exporting the data recorded by the app for manual processing (either the raw signals, or the little calm vs. active charts, or preferably both)?

An answer by a Muse employee would be appreciated :wink: (but if anyone in the community has some info, feel free to share it).

[I]Berkovich-Ohana, Aviva, Joseph Glicksohn, and Abraham Goldstein. “Mindfulness-induced changes in gamma band activity–implications for the default mode network, self-reference and attention.” Clinical Neurophysiology 123.4 (2012): 700-710.

Fell, Juergen, Nikolai Axmacher, and Sven Haupt. “From alpha to gamma: Electrophysiological correlates of meditation-related states of consciousness.”[I]Medical hypotheses[/I] 75.2 (2010): 218-224.[/I]


Also, what do the Easy, Normal, and Hard settings effect?


Home Improvement
Your posting is valueable and supportive …at your forum you describe the difference between calm and active states… i think different -different effect are to be build… these effects can be easy, normal,and hard… you have done an extra ordinary task…its my pleasure to see it… i am a new visitor at your forum but after saw your good forum i am fan of you… i see how you defined it in a very deeply manner…very well done… thanking you so much…for more information you can touch on this link : http://handyhomepros.com


I had some faith in getting my question answered on this post. However nobody replied… I guess my question is the same… What do you guys do in the app to detect if the user is calm or not? Why do we need the calibration phase before the exercise? What’s happening in the background during that phase?


Hey there,

The algorithm that the app uses to determine if your mind is calm or active is kind of our secret sauce, so I can’t say much about how it works. The calibration is necessary so that each time you use the app, we can build a model of your active mind at that time that the app can compare against.

The difficulty levels affect how dramatically the sounds respond to your brain signals.

Hope that helps a bit!


Hi Tom, thanks for your reply.
I do understand that the algorithm is a sensitive piece of information. However, any clues you guys could give here would be great, as most of us have limited or inexistent knowledge of the brain.

I’ll be more interested in clues like: which brain waves to focus on, and what sort of behaviour are we expecting when the user is focused? is it a peak of brain activity over a specific type of brainwave?

Any clue will be extremely helpful, and I’m sure whatever we (community) build, the more accurate we do it, the better for MUSE too as the users will rely more on the product too.
Thanks for your help.



Hey Andre,

Unfortunately my hands are a bit tied in that regard. The sauce wouldn’t be that secret if we provided hints on how it works. The information that is already publicly available on the choosemuse.com website and developer.choosemuse.com is all I can discuss. Sorry that I can’t be more specific!


Is there an email I could write to to make an official work collaboration request between Muse and my Client? can you point me into a non generic email? PM maybe :slight_smile:


Feel free to PM me!


As a therapist who utilizes neurofeedback clinically considering whether to recommend this to patients, I would like to know whether the frequencies trained by the muse app will result in overtraining a brain with particular proclivities within that band. An understanding of which frequencies are targeted would help make that decision as I am reluctant to recommend the product otherwise.


I had the same question because I want to develop an app that will connect to the head band.

Seems we can do some research on our own, there are thousands of studies and correlations.

But I would suggest we make an open source framework for investigation. I guess the method is quite simple, have a control group (non meditator) and monitor how they do with no meditation technique, not even focusing on breathing. Then detect the patterns seen in meditators. Some years ago I had some tests with another headband that measured symmetry as the main goal (left vs right), but I don’t think that is all we should look for. I noticed that different types of meditation do produce different effects in brainwaves. And there are many more things to look for different from just “Active vs Calm”. I have been able to get 70% “calm” in the app even while working on computer (my regular work with eyes open), and 95% even while having conceptual thinking (of course while being calm). I want to train the mind for stillness, so I wonder too about finding new patterns in the brainwaves. If you find any interesting studies please share.


Hello Tom,
I’m doing some of own research about muse (looking into What are calm seconds and calm points and how dose muse generate them [We know there is a calm multiplier of 3] and What are neutral seconds and neutral points and how dose muse generate them [We know there is a neutral multiplier of 1]) I see above you talk about a secret sauce that you can’t spill but I would love to talk to you about this for my project. In email or PM.