visual representations of sound.

[tester]

Any ideas how to emulate animated version of these ones?

Generally the whole idea refers to shapes that appear even when "stimulating" with single frequency; what confuses me is the fact, that single freq becomes irregular also on rounded/circular plates with centric location (or I'm missing something?). Best solution would be to define plate shapes, send the sound, and see the "mandalic" interactions happen (animated).

[Jay]

that's something i have thought about a few times but reckon it might be quite an undertaking to do it! I have played with a few software cymatic emulators like
http://www.falstad.com/mathphysics.html and http://www.resistance2010.com/page/cymatic-software

Fastlad has java source code for download on his site! it looks complex to me

it would look great though! Trogg! hee hee

[tester]

I bet it can be simplified.
Is this second app doing animations (i.e. live update) or only pictures?

[Jay]

I think it is just pictures on the second one as it takes a few secs to update the image! like as if it renders a still in the background then shows it! fastlads ones are much more in line with the real thing!

have you ever set one up? i do it with an amp and old hifi speaker and a metal tray filled with water and cornflower/ cornstarch and a little food dye! its great fun!

[tester]

No, not yet. Recently I was wondering if it will work with so called "vibration speaker" (bought for different kind of experiments), but I failed with that one; they make surfaces vibrate, but have to small power to vibrate something on them.

[trogluddite]

VERY tricky - the modes of vibration and resonances of two dimensional objects are extremely complex.
For a simple ~one-dimensional string each mode of vibration is (approximately) an exact harmonic, which can be calculated from string length, mass per unit length and stiffness. But two dimensional objects have resonances which are not related in a simple harmonic series, and which are free to propagate in any direction.
You can see on this Wiki page some animations (bottom of the page) of several of the most simple vibration patterns for a typical circular drum skin - but in reality, you would get a large number of those patterns superimposed. The medium on the vibrating plate then gathers in the areas of least motion.
This is getting well into the realms of physical modelling and finite element analysis, which as the reams of maths on that web page show are incredibly complex - orders of magnitude more difficult than, for example, FFT's and the like.
That will be why Jay's example is not a real-time animation - the skin must be divided into sufficient "cells" to give good resolution to the modelling - that's a huge array of points, to each of which all of that gobbledygook maths must be applied.

[tester]

Again - I believe it can be simplified, I just don't know how.

My impression is, that what that fellow did with framework programming (as far I remember correctly right now?) - is just slow because it's unoptimized "mainstream"; it's like making general apps in FS/SM (also - it's like playing with FFT; while it works horrible in FS - I work with a software, that creates high resolution - 132kpts and more- in "real time", i.e. acceptable speed). I suppose, there are some "tricks" that can be used instead of too real modeling.

One thing that could be done - is to use it for green layer (if sounds are not complex; for example sine frequencies) instead of FFT'ing blue. But I guess this is not the direction. But it would be interesting. Various "materials" and plate shapes could be tested/visualized. Right now - beyond my skills and knowledge.

[tester]

Hi Jay. Just out of curiosity I purchased this "Software Tonoscope" ($35), and... it's the worst spent money I had in last years. Bugs, unfinished routines and crashing, crashing, again crashing, (almost) whatever you try do.

So I think the question to emulate such patterns is rather open.