Beginner EEG setup questions


Hi all

I have been looking at the different EEG setups available for home use and am thinking of putting together an UltraCortex Mk3 with a Ganglion board. The purpose for this project is to experiment with different meditation techniques to see how they affect my alpha, beta, theta and delta brain waves.

I want to avoid anything with a chin strap because one of the things I want to do is see how my brainwaves are affected when playing the Native American flute before/during meditation.

I’m thinking of putting together an UltraCortex Mk3 with a Ganglion board and using the OpenBCI GUI. Before I jump headlong into this, I was hoping to get some feedback and see if the UltraCortex Mk3 and Ganglion board would be suitable or if there might be a simpler setup that would give the same results?

One thing that I would like to find out is how many channels I would need to monitor alpha, beta, theta and delta brainwaves? Would there be an advantage to having 8 channels over 4, or even having 16 channels? Another thing is, where would be the best points on the scalp to connect the electrodes?



Welcome to the forum @backspace. I haven’t used the UltraCortex, but if you’re looking for a headset that doesn’t have a chin strap and allows you to position your electrodes in a variety of locations, then it’s probably a good choice.

As far as I know it’s possible to monitor the various frequency bands (alpha, beta, etc.) even with only one or two channels. The advantage of more channels is that you can get a better picture of what’s happening in various parts of your brain at the same time.

Many devices that are used for neurofeedback have only a few channels (such as many Pocket Neurobics devices, and the Muse headset for example).

I don’t have any recommendations regarding electrode placement for meditation, though it’s something I would also like to find out more about.

Regarding software, the OpenBCI GUI is good for basic, real-time visualization. You could also check out the open source Brainbay or Neuromore, both of which allow more recording and visualization options than the OpenBCI GUI.

Muscle motion, especially in the head, can produce electrical artefacts that will be mixed up with your EEG signal. Getting a good recording while playing a flute might be tricky because of this. You might be able to do some processing of the signal to remove the muscle artefacts, but this would be a job for more powerful software than the GUI.

@wjcroft is the reigning expert on this topic, and might be able to give you some good suggestions if he’s around.

Good luck with your project!



Adam and Backspace, hi.

There is a ton of research available on EEG correlates of meditation, dive in!



I don’t want to dissuade you from OpenBCI, but also checkout Muse. A lot of folks using it as a meditation aid,



Thanks Adam and William. I see I have quite a bit more reading to do before I start shopping.

Thanks a lot for the info and the links. I’ll go away and digest as much as I can. I’m sure I’ll be back with more questions soon.


Thanks for the replies William.

The difficulty I have on this topic when looking at, for example, a Google search is that there’s so much information available. I think I’ve read at least a dozen papers about it, including review articles, but I’m still at a loss to answer apparently simple questions like “where do I put the electrodes to see some effect of meditating?”. Maybe this is because there isn’t a simple answer and it depends on the type of meditation, the person, etc. Still, I think it would be great to find a summary that pulled together conclusions from the literature in a DIY-friendly fashion without over-simplifying. Perhaps it’s a topic for a future post on

The NeuroMeditation Institute has a nice breakdown of meditation types and NFB training targets, which is a good start. It’s a bit short on details, though.



Hello hello folks,

I did use Emotiv for a while (with the EEG SDK) and switch to openBCI, been using both for brainwave detection for couple of years. Recently I receive my Gaglion boards too.

First , UltraCortex and the chin strap is not compulsory in my experience. The gal is keeping the sensors and I had good signal with gal only (not a lot of mobility thou), in my early prototypes I actually used just a winter hat to keep the sensors fixed. Later I made my own velcro headband, but I’ve also seen similar for $20 or so on the web. UltraCortex, I think, is giving you the advantages of using moisture instead of gal as well as the easier sensors placements, but it was too expensive for my case.

Most of the functionality from openBCI GUI for the Cyton board is not yet ported to Ganglion, but it looks promising. You would need to steam the channels input and do the filtering yourself, not long and trivial processes but ganglion just shipped and there is not much out, yet. NeuromoreStudio is great solution in general and offers a graph interface with bunch of examples and OSC, but no support for Ganglion yet and some strange pricing options, which I haven’t really know about since I used previous open betas. openBCI is providing great documentation and software libraries, like NodeJS and Processing, which is just lovely, but the Ganglion GUI is …in the process… Dunno when the full functionality is coming, but as the name of the product suggest, ‘open’ means you can easily do stuff if you want.

Brainwaves detection with fewer channels is indeed somewhat possible, but I don’t believe it will be good representation of brainwaves with only one channel. After all, one sensor means only one part of the brain so you cannot really make an objective picture. I did have the daisy module for a while, so I made tests with both 8 and 16 channels and was even happy with as low as 6 channels for my project. To be honest you can go with 2 sensors if you focus on frontal cortex for example and Alpha-Beta waves, but I’m not sure how far research is with association of brainwaves and parts of the brain when it comes to meditation. The 4 sensors however go nicely in front and the back, which is somewhat complete chain. Here it comes the tricky part in my perception, by having 4 sensors you can monitor signals on top or the edges of the head and some people do have different personality related to which parts of the brain are used. So maybe we need to place sensor slightly different locations depending on the person and such adjustment would be possible, but less needed with more sensors. The 16 sensors setup allow us more information from more regions of the brain, hence higher resolution brainwave readings and the possibility to discard channels without effecting much. This week I was playing around with 3 sensors on Ganglion because one was not active and I didn’t fix it for the sake of experimentation and while it was acceptable case it was very frustrating to have some part of the brain data completely in the dark… so yeah, any number of channels would allow you to define present brainwave, but the more sensors in your chain you have, the better you would be able to define the dominant brainwave and the given brain state. Probably 4 and 8 are the commercial minimum because it does represent somewhat complete net on the head.

The classification of the meditation in Emotiv (which is different suit than the raw EEG/brainwaves approach) is done by sampling many buddhist monks. I did not manage to record high meditation activities in my previous tests, but in theory you can teach your own system to classy mediation, probably just as good with openBCI and large enough valid meditation sample. Emotiv does have (sort of) learning abilities, as the device gets more accuracy over time for a given person (profiling), which is sort of cool. With openBCI the provided raw EEG signal, needs to filter and to classify meditation.

To make the long story short, openBCI does offer sort of barebones products with better quality of signal, documentation of the software, access to your EEG data and (most importantly) updates. Emotiv and Muse (probably others too), however does offer more ‘commercial’ ready product, where you can get classified data to dashboard and recording it, but somewhat limited access to the raw EEG data. I would choose openBCI any day, but you cannot easily ignore the experience completeness of emotiv and other commercial package. I’m not sure how non-developers make the choice, as it’s not easy.


Hello @AdamM , I’m totally newbie so probably I won’t be able to follow up yet, but could you please describe this processing? Do you write code for this, or is it just something like a configuration at a program like neuromore etc?
Thank you!


Good question @celtic_harp. I’m not really qualified to answer, since I have almost no experience with processing to remove artifacts. I know that various EEG analysis software suites have some of the basic tools that are required in order to do artifact rejection (at least OpenViBE), but I haven’t heard of any with ready-to-go filters for removing muscle artifacts.

Some artifacts can be removed relatively simply by filtering out a specific frequency, or removing sections of recording that go above some amplitude threshold. In cases where there are not simple features that distinguish artifacts from the EEG signal, more complicated techniques like Independent Component Analysis (ICA) are often used.

For a very brief overview, check out this Q&A on

Probably the first step would be to get a recording of the EEG in question and see what type of artifacts were present in the signal, and whether the artifacts would impact the intended application of the recording.

Googling something like EEG muscle artifact rejection gives a wealth of info on the topic, although most of it is quite technical and a long way from simply turning on a filter in a program GUI.

Here’s an interesting paper that is probably relevant to @backspace’s plan of recording EEG while playing a flute: Removal of muscle artifacts from EEG recordings of spoken language production. (This appears to be a draft version of the paper. The published version is behind a paywall, though it might be available through sci-hub.)

Good luck with your research and experiments!


Great! Thank you for your explanatory reply.
So I guess when speaking is not needed in an experiment you must forbid it at the time of the EEG recording to avoid EEG muscle artifacts.