AudioShake

AudioShake offers AI-based sound separation that decomposes existing audio recordings into component stems, including vocals, instrumental, drums, bass and other instruments, as well as dialogue, music and effects tracks. The technology works on material that was never multi-tracked, and also supports multi-speaker separation, isolating individual voices even when speech overlaps. Alongside separation, the company provides automated lyric transcription with word-by-word alignment of lyrics to the recording.

The platform is applied across several workflows: mixing and mastering, including creation of Dolby Atmos and Sony 360 immersive mixes and mastering of live recordings where original stems are missing; localization and captioning, where clean dialogue stems are extracted to improve ASR transcription while retaining original music and effects for the localized output; interactive audio for gaming, social media and fitness, where songs can adapt in real time to user movement or gameplay; sync licensing, where instrumentals are generated for pitches; audio analysis, where dialogue and music separation is used to clean input files before metadata analysis, sonic branding or content recognition; A/V editing tasks such as removing copyrighted music, cleaning noisy broadcasts and splitting speakers into separate streams; and copyright compliance, where isolation of recording and composition elements supports sample clearance, royalty allocation, ownership verification and dispute resolution.

Business model

AudioShake sells access to its audio separation and lyric transcription technology to media and music organizations, offering a self-serve free trial alongside a direct sales channel for enterprise customers, and describes low-cost automated lyric transcription as a service offering.

Traction

Stems produced with the technology have been used on catalog recordings from artists including the Jackson 5, Nina Simone and Whitney Houston, while AudioShake instrumentals have appeared in Disney and Netflix trailers and commercials for brands including Oreo and Taco Bell. Customers report transcription accuracy improvements of 25% or more when using extracted dialogue stems. Disney Music Group describes itself as an early adopter, and AI Media has partnered to isolate and clean sports commentary in real time for AI-generated translation via its LEXI Voice product.