AudioAlign is a research tool for the semi-automatic and automatic synchronization of audio and video recordings that have either been recorded in parallel at the same event or contain the same aural information, even if the recordings suffer from time drift. It uses audio fingerprinting to determine overlapping intervals in recordings and refines them with dynamic time warping. The tool can be used to synchronize hundreds of clips and offers an interactive timeline for manual inspection and refinement. The synchronized result can be exported as EDL (Sony Vegas) for post-production.
I originally developed it for my master’s thesis in 2010, later updated it for my doctoral research, and published it as open-source software. Its audio processing layer is available as a separate library called Aurio.
