/muse/dsp/elements/beta doesn't show anything I'm muse-lab with --dsp mode


#1

Hi, I am trying to visualise alpha beta theta and delta brainwaves. I added the --dsp flag and when I use oscdump with | grep /muse/dsp/elements/beta it does show values which are changing. But when i go to muse-lab it just shows a flatline. But I can see raw EEG data. Any thoughts?


#2

Hi there,
Just to clarify do you see any of delta, theta, alpha or beta changing in muse lab? The DSP will give constants or not a numbers if you are not receiving good data signal from Muse. Each channel will have it’s own values for delta, theta, alpha and beta and they will all look flat without a [B]quality EEG[/B] input.

You can also plot the [B]frequency spectrum[/B] in a [B]stationary[/B] line graph to be certain they are being sent. If you plot a stationary line graph in muse-lab and then plot the [B]/muse/dsp/elements/raw_fft0[/B], 1, 2 and 3 it will show a live fft of each channel. There will be many frequencies plotted her from 1-110Hz (0-110Hz are defined by the bins between 1-129 excluding 0Hz. The order of frequency bands, loosely defined, is delta(0.5-4), theta(4-8), alpha(8-12), beta(12-30), gamma(30+). Feel free to look closer into the frequency band specifications.

Related to your question though the [B]most important [/B]thing to check is for [B]quality EEG[/B] into Muse-IO. Noisy EEG will not produce any dsp frequency outputs.


#3

Thanks for getting back to me!
The noise in Muse-io is 0.0% on all channels, what i dont understand two of the channels are green whereas the rest are in red. Sometimes i do see noise but just 0.9% I don’t know if that is enough to disrupt the DSP output. But i think i am getting a quality EEG output, because if I plot the EEG and do the eye blink test or clench my jaw, i do see that depicted in the graph.

I start the muse-io in udp mode on port 5000, when I then use oscdump 5000 | grep /muse/dsp/elements/beta do see values which are changing, from which I assumed that it was working. If i try other channels they will also show data just different values.

I will try to plot the frequency spectrum. I will let you know what i come up with.


#4

Hi,
Sorry I just reread your original post and I misunderstood your original issue. The plot on Muse-Lab was [B]flat[/B], but the values are changing in osc output.

This is most likely due to the fact that your zoom levels on Muse-Lab are too far out. The dsp values for frequency bands are ranging from 0 to 1. The default plot size is on the scale of -10,000 to 10,000. Making the scale of 0 to 1 look very flat. Try adjusting your plot scale to properly plot the delta, theta, alpha and beta values.


#5

Thanks Farough, by changing the scale I can see the waves changing. One more question does a higher value mean a higher voltage? So would .5 beta and .2 alpha mean my beta waves are stronger and alpha weaker and indicating that I am more alert than calm. ( I know this is very much oversimplified, but just for me to get an idea)


#6

The voltage being sensed is AC and alpha/beta are bands within a spectrum of lower frequencies we are looking at >1hz, so a higher value means more power in those frequencies of the AC (alpha/beta) but that does not mean not more or less voltage. It’s like if a speaker was humming a constant tone, it could be loud or quiet, and the pitch doesn’t relate to the volume.


#7

Thanks for replying MattC the example with the speaker explains it well. I think I confused power with voltage. So to continue the speaker analogy, if the alpha/beta/etc bands are the tones, the values that are displayed is the volume? So when the alpha band jumps from .3 to point .5 it would mean it had gotten louder?


#8

With our DSP outputs it means that one has gotten louder relative to it’s recent history. Everyone’s skin has a different resistance and the impedance to the ADC can vary. So when it jumps, it means the frequencies in the alpha band have more power then they recently did. To relate it to sound the volume of the tone that is “alpha” has almost doubled it’s past “loudness” when it jumps from .3 to .5.


#9

Thanks MattC.