ANIMUS / Module 01
VoxDNA
Extract reviewable voice identity from local media.
Workflow
Review stays in the loop.
Import raw media, review speaker evidence, recover uncertain fragments, and export only the material you have checked.
Interface details may continue to evolve during beta. The live app reflects the current build.
Evidence Review
Clustered voice evidence, prepared for human lock.
VoxDNA turns raw local media into speaker pools and reviewable fragments. The operator confirms transcript, speaker, emotion, boundary, and export state before evidence leaves the machine.
Cluster Review
Group fragments into speaker pools, inspect candidates, and correct identity labels.
Fragment Detail
Review transcript, emotion, and boundaries inside the fragment before it is trusted.
Human Lock
Confirm what becomes reviewed evidence before it leaves the local project.
Outputs
Your raw media is not the product.
Raw media stays local. Reviewed evidence leaves only when exported.
Reviewed clips
Speaker-aware fragments prepared for downstream use.
Subtitles
SRT and ASS from reviewed transcript evidence.
Structured data
CSV and JSON for sorting, filtering, and future layers.
Timeline Bridge
AAF handoff for external timeline review where supported.
Dual Interface
One project state. Two operating surfaces.
Guided review and Professional Timeline read and write the same local evidence model.
Default entry
Clustered Evidence Field
Clustered review for operators who want identity grouping, verification, and export without timeline navigation.
Precision surface
Professional Timeline
Timeline review for operators who prefer tracks, boundaries, clip-level editing, and media handoff.