It's constantly fascinating me lately how the consonance or dissonance of given intervals depends so much on the timbre of the instruments. I did something last weekend that really surprised me, even though I should have known it: - first, grab a guitar, bass, piano, etc and play a major second harmony on it (somewhere around middle c -- if it's the bass, play up high - may require detuning one string if you can't do the stretch). It's not a dissonant interval, but it's not exactly consonant either. There's a lot of warbling going on. That's nothing new. Next: - get a set of tuning forks - grab two that are a major second apart, eg a G and an A - give 'em both a whack, and hold them near one ear You'll hear the same warbling. Finally: - take those same two tuning forks - give 'em a whack, then hold one up to each ear No more warbling! Just a wonderful sounding harmony! The warbling is caused by the constructive/destructive interference between overtones of each note that don't quite line up exactly. The interference happens in the air, as the waves meet. When you hold them to separate ears, the sounds don't add together in the air (presumably they're not that loud), your brain just hears both at the same time. This is really cool for a number of reasons: - you could alter the perception of consonance in a recording meant to be heard through headphones by splitting harmonies up to the two channels (or putting them both on one channel) - the really cool one: this deomonstrates that, whatever is going on inside your head from the ears inwards, is not representing pitches directly as an oscillation. If it was, at some point the signals from the two ears would reach the same area of the brain, and the same constructive/destructive interference that occurs in the air would occur electrochemically. Instead, it's like the information about the sounds just gets superimposed. One ear hears the G harmonic series, the other hears the A series, and the result is just all of the above. Science already knows this (I think the sound gets divided up into hundreds of "EQ bands" to use an analogy), but it can be demonstrated with only some tuning forks and a little knowledge of wave theory. No electrodes or PET scans needed!