Five great auditory illusions (http://www.newscientist.com/channel/being-human/dn13355-music-special-five-great-auditory-illusions-.html?feedId=online-news_rss20)

This was fun :)

Yes, it was. Thanks, Michael.

I haven't tried "Phantom Melodies" because the file momentarily crashed my browser (Firefox) -oh, well. :cheers:

That was pretty cool.

Cheers

The throat singing high voice in Tuvan music (I couldn't listen to the link for "quintana" because the bandwidth limit was exceeded) is not an auditory illusion - it's actually being produced. All the rest of the vocal sound can be cancelled (and the best syggit singers can all but suppress it too inaudibility) and that high tone will still be there. It can also be changed - used in melody - without changing the ground tone.

Here's two examples of throat singing:

http://www.youtube.com/watch?v=DY1pcEtHI_w
http://www.youtube.com/watch?v=6vkejOp9wOc

Wow that pretty cool. I've never heard of throat singing before.

I think the text explaining the Barber moving around your head as due to the intensity in one ear being greater than the other is part of the effect, but much more important in normal hearing are the the phase delays. This should be clear to you if, for example, you heard a distant car horn blow. I.e you easily locate the direction from which it comes, despite fact both ears have nearly identical sound intensity acting on them. ("Head shaddowing" is the only significate intensity difference, and it is nearly nonexistent for source near "front ahead," but when source is not exactly "in front," the phase difference is essentially independant of the source difference.*)

I believe there are ear phone studies, which clearly show that the phase differences dominate. I.e. if phase difference indicates source is at left, you will perceive the source at the left, even if the intensity in right ear is quite significantly greater. You can easily prove this also: stuff some cotton in one ear to reduce the intensity (but not the phase difference) and the perceived angular location of the source will not significatly change.

The phase differences are zero for source either straight infront or directly behind, as are the intensity difference, but you can usually still tell which alternative has the source. This has to do with the reflections in the outter ear, mainly the higher frequency components. (You will have great difficulty if head is constrained against even slight turning and only wavelengths large compared to you ear are present in telling "source at front" from "source at rear.)" When the source is at your left or right, your angular error to localize it is maximal. (To make a given phase difference, more angular shift is required.) Note that the phase difference are NOT a significant function of how far away a distant source* is, but the intensity difference are mainly due to "head shaddow" effects when the distance is great.
---------------------
*The sound waves coming from it are essentially a plane wave advancing over you.

The audio illusion I like best is the "constantly rising pitch" - I do not have reference, but some one good at searching should be able to find. Here is how it is achieved**:

A set of notes (uniformly spaced hertz gaps, as I recall) are all moving higher in pitch but their amplitude is under a "bell curve." I. e. each newly created note at the longer wavelength end of the bell curve is very weak but grows in intensity for a while then begins to die out. What you are most aware of is the dozen or so stronger ones and they are all rising in pitch with a constant hertz gap between each. For example if the "hertz gap" is 100Hz, then at "T = 0" when one, note is at 1000Hz, it will soon (at T = 1) be at 1100Hz and then exactly the same set of of notes with the same set of relative intensities as at T = 0 is again restored. I.e. what you hear at T = 1 is exactly the same as what you heard at T = 0, yet the pitch of all sounds you hear is constantly rising!

For those who already know the visual "water fall effect" this auditory perception resembles it but the mechanism has nothing to do with fatigueing certain neurons of the brain (the motion detectors) - Called the waterfall effect as a waterfall is a natural scene which is constantly in motion but never moving position. Location processing neurons are in different brain location and do not fatigue (because of the constant and very fine "jitter" of the eyes. - Parts of a retinal after image can drop out of perception by fatigue as the image is fixed on retina so it does not have this "jitter.")
--------------------
**Perhaps someone good with computer programming will build a demonstation (Amplitude under a "tent curve" should work just fine and be easier to compute than under a "bell curve")

With the auditory illusions talked about, and how music in particular is a way to demonstrate it, it's partly what we hear and partly what we expect to hear.

As with the fast/slow piano sequence and a melody appearing, it's to do with the patterns our auditory centres "settle" on.
David Bowie's Ashes to Ashes, which is played "slowly", has a bunch of illusory notes in the intro - the first bar sounds like a sequence of 4 high notes, but there are only 3 played, then two sequences of 3 notes, but only 2 are played each time.

The virtual barber shop was fantastic!