AudioTexture is based on concatenative synthesis which analyzes automatically the sound signal and decomposes it into adaptively defined units (not equal in size, so not the usual granular synthesis) for resynthesis. It is equipped with semantic descriptors for intuitive control to generate many variations from a given sound sample. The main features are:
- Automatic unit analysis.
- Position and Descriptor control.
- Range control for randomization.
- Multichannel support.
AudioTexture comes with 3 descriptor controls in additional to the position control. This is convenient to create various unit sequences based on perceptually meaningful descriptors:
- Energy: Instantaneous energy.
- Noisiness: The amount of high-frequency content.
- Brightness: The spectral energy center, higher the value brighter the sound.