Getting data for music visualization

Question

Hi!

I had an idea of a music visualization project (like those in winamp and other players), but as I am completely new to this, I don't know how to get the data (frequency spectrum, loudness, etc) from a music file. So, what you would recommend to get the data? There are some libraries for programming languages (i know some c++, python) or programs (i am using linux) that do this?

(The visualization won't be done in real-time. The sound and video will be combined together later).

Thanks in advance,

-skazhy

Answer 1

Your first port of call should be Processing, which is a Java based language with a simplified syntax, limitless visual/graphics capabilities and good support for audio.

You could also try packages such as SuperCollider and PureData, both of which run on Linux. PureData involves no programming, as such, at all -- rather it is a 'dataflow' programming language, aka point-and-click. Still, many interesting results are possible. SuperCollider is a powerful language aimed more at audio programmers and composers, but probably has the best feature extraction (ie, audio analysis) options. In the past, I have created visualisations by extracting the audio data in SuperCollider, and sent it via OpenSoundControl to Processing. This would be involved, but potentially lots of fun..

How you extract frequency spectrum, loudness, etc depends on the feature you are extracting and the platform you are using. Generally, it is not too difficult to do: your first step should be to check the platform documentation.

Hope that helps