VASILISA - 16-fold scalable GPU-accelerated (WebGL) wavevet video codec

Four levels of spacial scalability based on wavelet transform

Vasilisa is based on Schroedinger video codec and boasts four levels of spacial scalability. That means a single video file contains four various spacial versions of the same video each version twice as big dimensions-wise. These versions are not separate with higher dimensions ones taking advantage of the information in the lower levels thus saving lots of disk space. Our HTTP server module has simply to copy the necessary chunks of video data form the file to the bitstream without any heavy re-encoding.

Four levels of temporal scalability

Vasilisa has four temporal levels of scalability out of the box, meaning each video file encoded with it can be played at four different frame rates: full, one half, one fourth or one eighth of the full frame rate. These are independent from spacial level that is you have 4x4=16 (sixteen!) different versions of the save video in a single file and you can choose which one to get by, for example, sending a GET parameter to the HTTP server.

Ubiquitous support and extensibility

Vasilisa's decoder is written in WebGL, Wasm and JavaScript which are supported by any modern browser meaning the video can be played on any device! And we can add any new feature without the need to wait for the browser vendors to implement it! Full control.

Alpha channel (transparency) support

Vasilisa supports alpha channel meaning you can play fancy transparent videos in any web browser!

The tech

Vasilisa's decoder features RANS binary entropy decoder in WASM, that follows by motion compensation in wavelet domain using complele to overcolete wavelet transform and inverse wavelet transform - all in WebGL( GPU ).

Interested in having your own video codec that is supported everywhere ?

Get in touch now