Home

This documentation contains very in-depth explanation of Sushi's inner workings, reasoning behind some of the design decisions and detailed description of command-line arguments.

The general algorithm is very simple:

1. Sushi prepares everything it needs during the demuxing stage
2. Shifts for all events are calculated based on audio search
3. Some postprocessing is applied to cover possible audio search errors

You can also check out some usage examples and read about common errors.