Hallo everybody,

I just found an interesting website:

http://www.schulphysik.de/java/physlet/app...s/illusion.html

Sorry that it's in german.

Just press the buttons going from left to right. C1, D1, ... At the time you arrive at C2, you'll think the tone pitch increased steadily... but that's not the case: just press C1 again and you'll notice, that they C2 and C1 are the same.

Regards, fileman.

Maybe for a music listener but not for an amateur musician

What?? That's ridiculous! Is this page for real? Did they even bother listening to it?

Nope. It pretty obviously drops an octave. It follows a certain scale rule, where an octave drop for the ending note sounds natural.

It tries to fool the casual listener with some harmonics in addition to the base frequency. But a regular synths player is harder to deceive

Yup. You are absolutely correct. First and last tones on that scale are the same Another ear fooled.. ... .. not!

This reminds me of a loooong, long time ago when ppl actually used to fool around with short-wave radios (you know, before the Internet and global-webcasting blew it away).

Some would tune across various bands picking up all sorts of carrier-waves with audible beat-frequencies skipping across the ionosphere. One such audio effect made it sound like there was a rich harmonic-tone with a continuous RISE in pitch. These lush 'harmonic tones' sweeping ever higher in tonal-pitch and would go on and on... never seeming to end. Yet the actual sonic 'scale' of the sound stayed relatively the same over and over!

Listerners to the Short-wave band radio producing this effect would ask. "Why doesn't that sweeping-scaled sound rise up to the next octave and continue upwards till outside of the audible (or speaker) range?" Well, there lied the trick: while the 'main' tone pitch was perceptively /rising/ upwards continuously, all the while there was also a 'filling-up' of the missing lower-tones continually swelling beneath it and rising in the same incremental pattern as the primary tone.

After a certain point in time, those NEW tones would in effect 'take-over' the tonal sweep and this would create a continuous pattern which fooled the ear into thinking the scale was 'never ending.' I believe this is the exact sonic-effect that Web Site is trying to achieve with your computer's own cheap tone-generators. haha.

I know. What does this all have to do with the price of eggs. Nothing. But this is a slow morning and such memories were floating about.

SJ

Yes, the thing indeed doesn't work perfectly, but it is an ingenious little idea nonetheless. Unfortunately, you wouldn't know what the "idea" behind this thing is, unless you understand German . Well, I do, so here you go:
What they are trying to do is basically this: You start off with a tone that contains a base frequency and a lot of overtones, all at the same amplitude, for example 13.75, 27.5, 55, 110, 220, 440, 880, 1760, 3520, 7040, and 14080 Hz (it is important to use a lot of frequencies). Then you increase the pitch step-by-step, i.e. you multiply all the frequencies by 2^(1/12). You repeat this twelve times, and then you arrive at basically the same tone you started with. 13.75 Hz has become 27.5 Hz, 27.5 Hz in turn has become 55 Hz, and so on. The only differences in the resulting tone are, the lowest frequency is missing (which you won't notice if you started of with a sufficiently low base frequency) and the highest frequency has gradually increased to 28160 Hz (for the example above). This frequency, of course, is far too high to be hearable. So the final tone will actually sound exactly the same as the first, although the only thing you did is increase the frequency continually. This might appear paradoxical, and that is the reason why the effect can be somewhat surprising at first listen. They don't try to "fool the casual listener". The thing really just does exactly what it says it does, it steadily increases the pitch. No trickery involved.
However, on that webpage, they probably didn't use enough overtones, so you can definitely notice that something is missing when you switch from H1 to C2 (well, at least if your high frequency hearing is ok). So I have made some files myself, employing the same principle, but using more frequencies and a sine sweep instead of single tones. They are a bit more disturbing than the example on the webpage. If anyone's interested I can upload them.

edit: Ah well, looks like SoundJudgement came in first.

I've uploaded a set of samples here. I hope it does a better job.

http://www.seeingwithsound.com/javoice.htm