Skip to content

1.4 Editing: audio and arrangement

Deepen the arrangement and audio-engineering surface so products can cover multicam shoots and serious audio cleanup without leaving the SDK.

#ItemPrioritySizeNotes
1.4-1Multicam editingP1LSync groups (waveform/timecode alignment), angle switching and a multicam playback mode, built on the existing time-mapping and track model.
1.4-2Per-track audio effect chainsP1MEQ, compressor and de-noiser as pluggable per-track processing nodes beyond the current fades/ducking/master settings.
1.4-3Voice enhancement and de-noisingP1MReal-time denoise and dereverberation as a node type in the effect chain; provider-based so the on-device/cloud decision stays open.
1.4-4Project templatesP2MParameterized Project instantiation: a Project acts as a template with editable template parameters.
1.4-5Advanced clip waveform/metadataP2SRicher per-clip analysis metadata surfaced to the timeline and preview.
  • Multicam sync produces sample-accurate alignment with validated timecode handling; angle switches are transactional.
  • Effect-chain nodes share one processing interface, bounded memory and cancellation semantics with the rest of the audio pipeline.
  • Template instantiation is deterministic: the same template and parameters always yield the same Project.
  • All new entities follow the complete execution path (builders, validation, migration, IR, preview, export, types, docs).
  • 1.4-1 depends on 1.2 analysis outputs for waveform/timecode sync and on the Render IR structure.
  • Audio effect graphs grow CPU cost; the 1.1 resource-budget and soak work must be in place.
  • Template parameters may require an additive schema extension; keep it optional and validated.