“Hand claps are a relatively primitive conveyor of sonic information, yet they are widely applied for different purposes.”

Primitive

though they may be, given just a simple audio recording of a single

handclap, is it possible to determine the hand-configuration of the

clapper via a computer-assisted digital analysis of the sound?

Researchers Antti Jylhä and Cumhur Erkut at the Laboratory of Acoustics and Audio Signal Processing, Helsinki University of Technology, TKK, Espoo, Finland reveal that, yes, it is.

Their paper: INFERRING THE HAND CONFIGURATION FROM HAND CLAPPING SOUNDS (published in Proc. of the 11th Int. Conference on Digital Audio Effects (DAFx-08), Espoo, Finland, September 1-4, 2008) describes their experiments in detail.

Using a combination of FFT filtering and the Steiglitz-McBride algorithm

on 40 recordings of claps (two clappers, 20 claps each), the team have

discovered eight distinct classes of clap. Named as: P1, P2, P3, A1, A2,

A3, A1+ and A1-  The results showing too

that “… it is possible to make inference of the hand configuration of a

clapper given the resulting sound.“

The paper also points to the possibility of more-advanced clap analysis

techniques that may be of interest to security researchers – but which

could perhaps be of concern to privacy-rights groups. Clapper Recognition.

In the future, might it be possible to electronically identify the

clap-signature of a particular individual embedded within a crowd of

clappers? “The results also suggest that the claps of individual

clappers may incorporate systematic differences from other people’s

claps …” say the research team.

SOFTWARE: The Helsinki University

of Technology Laboratory of Acoustics and Audio Signal Processing has

created a software package called ClaPD – one of the very few,

perhaps the only, cross-platform freeware fake-applause generation

software.

A video describing the software can be viewed (and heard) here (.wmv format). And ClaPD can be downloaded here: (Windows, Linux and OSX 10.5)