Extracting Binaural Beats from audio

How do you “extract” a binaural beat from any selected audio.
I have both downloaded and CD versions of music with binaural beats included,
however I wish to remove “those” and insert my own created beats.
Thanks so much for the help.

Use the “Notch Filter” to remove the tones that produce the binaural beats.http://manual.audacityteam.org/o/man/notch_filter.html

Thank you Steve for your input, I will try this.

However, please allow me to explain a bit about what is happening.

Binaural beats occur at a “set” frequency, let’s just use 300hz for an example, this is the “base” frequency.
Which of course you can use any frequency within the audible range (human) 20hz-20khz.
Then a "beat frequency is applied, let’s use 10hz (within alpha brain wave frequency range).
The 10hz (beat) is divided by 2 and then added and subtracted to the 300hz (base) frequency
to get two channels 295hz and 305hz, which is now a binaural “beat” at 10hz for “brain entrainment”.

My dilemma is that:

  1. I do not know the “base” frequency used
  2. Nor do I know the “beat” frequency used
  3. Since there is an “offset” (beat) of frequencies between channels, how do I “notch” them both out?
  4. Also, most of the “music” has “pink noise” added to reduce to “irritability” of the binaural beat tones.

I guess what I really need is a way to “read” the frequency of each channel (L/R),
then subtract the two to get the “beat” frequency used within the audible spectrum.

Like maybe some kind of waveform analyzer or oscilloscope perhaps?
Does Audacity have (provide) such an add-on (plug-in)?
Not sure what it is that I need, so sorry for the technical misunderstanding.

Thanks so much for your help.

“brain entrainment”.

And with that one phrase, you slid gracefully into the region of Subliminal Enhancement.

Our objection to these systems is not that we can’t do them, it’s that we have no tests for them. We have very specific tests and procedures for the vocal readings by someone wanting to publish an audiobook. We even automated several of them.

Subliminal? Not a clue.

1. I do not know the “base” frequency used
2. Nor do I know the “beat” frequency used
3. Since there is an “offset” (beat) of frequencies between channels, how do I “notch” them both out?

We can do number 3…


There is one significant note about these systems. DO NOT use MP3 anywhere. MP3’s very reason for existence is to delete subliminal sounds. If you downloaded any of these works or clips via MP3, then they no longer have any of those tones and frequencies.

And there’s no good way to test for them even if they were there.


Don’t the downloaded and CD entrainments say what beat frequency they use?

You can use the Audio Track Dropdown Menu to split the stereo track. There is a Spectrogram or similar view in that menu. You did not say what Audacity version you have, so we can’t be specific.

Why not just start from scratch with a song, some pink noise and add your own beats to that? In other words why download and buy CD entrainments that seem to be at frequencies you don’t like?


Using the track Spectrogram view (see: http://manual.audacityteam.org/o/man/audio_track_dropdown_menu.html#spgram), if the “Window size” is set large enough in Preferences (see: http://manual.audacityteam.org/o/man/spectrograms_preferences.html) then the binaural tones should be visible.
In this example there are constant tones at 296 Hz and 304 Hz. The Window size is set at 16384. If you look carefully you can see a straight horizontal line at about 300 Hz. Zooming in closer (vertically) will show the binaural tone frequencies more accurately. You will then need to split the stereo track so that you can apply the required notch frequencies to each channel (see: http://manual.audacityteam.org/o/man/splitting_and_joining_stereo_tracks.html)

Thanks so much, I did know this, mp3 is the lousiest format to use. I download in CD quality .wav or .flac formats.

Thanks so much Gale, sorry for the lack of information.
My Audacity version is 2.1.0, Windows 7 Home Premium SP1

I’m currently trying out the spectrum function, really cool, have to play with the settings.

Here’s my dilemma, I wanted to know what “their” beats were, so I could “recreate” my own.
I really liked “their” background music and figured since I “already” purchased these,
maybe I could “extract” their beats and then “reinsert” my own.

Thanks so much, I did try this and it really works great.
I did have a question though, how did you know exactly that the “tones” where at 296hz and 304hz?
I expanded the waveform, as you suggested, however the side (left) bar does not breakdown that precisely.
Is there a setting (value range) that can be set on that left bar graph?

Did you left-click over the vertical scale to zoom in? You have to do that as well as setting a larger window size.

Steve “knew” because 300 Hz is the only place in the spectrogram where the left channel shows a bar at a slightly different frequency than the right channel.


I knew the exact frequencies because I made it as a demonstration :wink:
It’s not possible to see “exactly” what the frequencies are, but by using a large “window size” and zooming in very close on the vertical scale you can see to within a few Hz. I think that Audacity 2.1.2 (due for release very soon) allows you to zoom in a bit closer than previous versions - see here for the current release version: Audacity ® | Free, open source, cross-platform audio software for multi-track recording and editing. - register at the bottom of this page if you wish to be notified by email when it is released: http://web.audacityteam.org/)

When you have guessed the frequency as closely as possible, apply the notch filter to a short section and look to see if you have guessed correctly - by looking at the spectrogram you can see if you have slightly “missed” and can adjust the notch frequency as required (Ctrl + Z to undo).