Telos Alliance has added AudioShake source separation as a subscription module for Minnetonka AudioTools Server.
The module can split a finished mix into dialogue, music and effects, or generate a clean music-and-effects track. Telos lists dubbing, language replacement, dialogue enhancement, music rights changes, restoration and remixing as target workflows. The separation step can be chained with loudness, QC, transcode and delivery modules.
Why it matters
Many archives contain only finished masters, not the original multitrack sessions. That limits localisation and reuse because dialogue or licensed music cannot be changed cleanly. Automated separation can create a workable starting point at library scale.
Derived stems are not equivalent to original production tracks. Artefacts can become obvious after dialogue replacement or aggressive processing, and rights teams must not assume that technical separation grants permission to reuse components. A production workflow needs confidence thresholds, listening QC and a route to reject unsuitable material.
