AudioShake's LyricSync supports lyric transcription and alignment in over 40 languages. The pipeline handles non-English languages including those with character sets and phonological structures that challenge standard ASR systems trained primarily on English data.

Yes. AudioShake transcribes and aligns lyrics from both new releases and legacy catalog recordings, including mastered tracks where no original stems exist. Legacy catalog is a primary application — many older recordings were never manually transcribed and original session files are unavailable.

Lyric alignment assigns a precise timestamp to each word in a transcription, synchronising text to audio at word level — powering scrolling lyrics on streaming platforms, karaoke tracks, and music accessibility features. DSPs require time-synced lyrics in timed text formats. AudioShake's alignment output meets format requirements for major streaming platforms and supports lyric video production and catalog metadata enrichment.

LyricSync isolates the vocal stem from the full mix first, then applies AI transcription to the separated vocal track — producing more accurate results than transcribing from a mixed recording where instrumentation degrades speech recognition. The pipeline returns time-synced lyrics at word level, ready for DSP delivery, lyric video production, or catalog metadata enrichment.