Many people know about the Shepard-Risset Tone/Glissando, which is a tone that starts low and seems to continue ascending(or descending) when you listen to it, even though it actually is playing the same thing over and over again.

Wikipedia has a nice article on this.

However, I read that Risset also made another sound illusion: The so-called Risset Rhythm, a drum loop that seems to keep on going faster and faster (Wikipedia does have an example of this, but no explanation). However, I'm unable to find any explanation of how this illusion works.

Could somebody explain it to me please?


2 Answers 2


Sounds to me like exactly the same principle.

The first rhythm gets faster and faster until it becomes a blur of noise and is removed from the sound, but over the top of that is superimposed the same rhythm at half speed. While you're listening to the first rhythm get faster, the second does the same, and eventually becomes the main focus of attention. By the time it does so, it's probably got to the point where you can just about hear it again at half speed.

The same piece of audio can contain both the original rhythm and half-speed version, so gets looped to create the illusion.

As you notice the rhythm getting faster, listen out for the same features, but in half-time.

  • 1
    Yep, definitely seems to be the same principle. I can hear the same thing going on; in fact, I could concentrate and isolate the slower rhythms fairly easily in this recording: swiki.hfbk-hamburg.de:8888/MusicTechnology/uploads/826/… Nov 27, 2011 at 4:15
  • It sounds as if the volume of the rhythm is a bell curve function of it's closeness to "correct" speed. So, it fades in at approx half speed, reaches max volume at full speed, then fades out at approx double speed, always mixed in with its previous iteration.
    – slim
    Nov 28, 2011 at 9:59
  • Plus as the original gets faster and faster it gets shorter and hides (or blends) into the backbeat of the slower one so it's less and less audible. Not to mention that the faster sound exits as a kind of octave higher of the original snare accent.
    – user1306
    Jan 4, 2013 at 2:04

I think that the effect is actually created solely based on the tone played after each repeat, which is sort of like the glissando in that it increases in tone forever, but the actual rhythm never speeds up. You can test this theory by listening to the sound, then quickly turning the sound off and counting time in your head, then adding sound back in. You will find that your time counting matches up PERFECTLY with the audio upon re-adding the sound back in.

  • 2
    This is simply false. Try tapping simple crotchets along with the entire recording. You won't be able to, because what are crotchets at the beginning end up as 64ths at the end. Oct 31, 2016 at 23:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.