Parker Roan

Career software engineer and founder of Recastr.

Posts by Parker Roan

Music vs cinematic source separation for narration over songs

Two kinds of separator, given narration over songs with lyrics. The music model put the singer in the speech stem every time. The cinematic one mostly did not.

September 10, 2026 · 7 min read · separation, background-audio

Whisper, GPT-4o, Gemini, Grok or Azure: which transcribes best?

Eight speech-to-text models, ten runs each, the same eight clips. The most accurate, the cheapest and the one that can build a timed SRT are not the same model.

September 9, 2026 · 9 min read · transcription, benchmarks