Audio Cutup PD Patch

aka the Speech Finite State Machine

attachment:audiofsm_burroughs.png

Download

attachment:audiofsm.zip (new! version that supports multiple audio files)

attachment:speechfsm.zip (first version, single audio only)

To Use with your own audio file:

The sampleplay patch needs audio in a particular format, namely 44.1k/16bit (mono) as an AIFF file. (You can use the UNIX tool sox, or Mac users, the export function of QuickTime to save your audio in this format). Place the file in the same directory as the patches, and change the reference in the sampleplay object (see patch).

<!> In the "new" patch, the first word (burroughs in the example) is used as a kind of id. You need to type the same word then in the playfrag objects. In this way you can have multiple sampleplay objects for different audio files (use a different id for each).

<!> Remember the trick to saving your work in the speechfsm patch is to use the "tool" part of the patch (with the two number boxes) to figure out good start and duration values for your audio, and then to make a message + "playfrag" object, with the two numbers in the message (like "5.6 0.89"). These messages will be saved when you save your patch -- the contents of the numbers boxes will not!

Use the shift key to select fractional values in the number boxes.

The example pictured above shows a four fragment FSM with a single control to allow a change of direction. The patch, which uses the newer attachment:audiofsm.zip patches, can be downloaded here: attachment:audiofsm_burroughs.pd

Resources

• Brion Gysin resources on ubu