# Can you please explain "Scanned Synthesis"?

I just purchased (an amazing deal, pay what you want for) Scanned Synth by Humanoid Systems. I'm curious about scanned synthesis. I've read the Wikipedia page on it, but I'm wondering if someone could explain it a bit more clearly.

Cheers!

Eric

(source: humanoidsounds.co.uk)

This is how I picture it, and I may be wrong.

Imagine a piece of string attached from both sides, now add an LFO to make that string oscillate. The LFO doesn't have to be a simple wave (sine, square, saw, etc.) it has physics principles that allow it to have a user-defined life for a certain amount of time. Same as a guitar string, but with unlimited oscillating options. So, you can make the string be plucked at the center, and take X amount of time to die out, or that the string is very tight but a big plucking force is applied thus dying out quickly with a fast attack. Maybe this could be compared to a physics engine for string? And have a sort of ADSR type evolution?

The scanning part, I think is how the pitch is controlled. So if the above gets scanned periodically at 440 times per second, then the frequency will be 440Hz.

Now if you have control over the shape of the oscillation (timbre), and have a separate control over the speed of the scanning (pitch). You have a system that allows you to control the timbre and the pitch separately.

Basically, you can use this to build a string vibrating physics model.

UPDATE:

I found this video, check from 40 sec onwards, it explains well:

http://vimeo.com/13975651

[vimeo]13975651[/vimeo]

• Excellent explanation for the visual learner... Commented Jun 24, 2010 at 22:02
• thanks for that video add, Andrew... very clear and to the point. Commented Oct 25, 2010 at 14:52

Haven't heard the sound of it yet, looks interesting. Thanks for sharing...

Picture that comes to my mind is that there is an independent dynamic system to be used as the "material" for the sound synthesis, and the "movement" of this dynamic system is determined by some wave function...think of the surface of waving water...sample the "shape" of this surface at regular intervals to give shape to the generated sound according to some other function that determines the pitch and the timbre...

So it's just turning something into something else...would like to see this kind of system work with real-world objects such as waves, wind, or some other dynamic surface...or even real-time data.

I liked the way Andrew put it. I don't quite understand it but I found this article from cSounds to be a little more helpful than the wiki page. It also includes samples which is cool.

Definitely something interesting worth looking into though.

Hope this helps.

The Sound on Sound review of the product from back in 2007, also links this PDF sheet about Scanned Synthesis.

http://www.billverplank.com/ScannedSynthesis.PDF

alt text http://web11.twitpic.com/img/120152534-e5bc23a8df86c1dd9531af3c702c3ea3.4c2364e0-full.jpg

"Figure 1. Scanned Synthesis consists of 1) manipulating a dynamic system and 2) scanning out a wave-shape from along a path."