Wavelet Based High Resolution Sound Texture Synthesis
This paper describes the adaptation of Efros & Leung's pixel-based Image Texture Synthesis (ITS) to 1-D for Sound Texture Synthesis (STS). The goal is the creation of a long, dynamic, sound "texture" from a much shorter audio training example. The Dual-Tree Complex Wavelet Transform (DT-CWT) is used for optimization, to good effect. We define the concept of High Resolution Sound Texture Synthesis (HR-STS) as the texturing of high resolution, multi-channel sound recordings with retention of stereophonic effects. HR-STS is useful for installations, computer games, audio repair and low-bandwidth media devices. We test a variety of real-world training examples including ambient sounds, speech snippets and music. The resulting sound textures are plausible and varied without sounding "tiled"' from the training examples.
Click to purchase paper or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.