sxsphinxstack

Skills / Working / Create an audio delivery package

Working Video & audio skill

Create an audio delivery package

Create an audio delivery package with correct masters, stems, channel layouts, timing, loudness, metadata, rights, and verification.

Deliver files another operator can identify and verify without guessing.

When to use

  • Use for podcasts, video mixes, radio, broadcast, film, games, ads, localization, and archives.
  • Do not include unlicensed sources, private session material, or credentials in the delivery.

Procedure

  1. Obtain the current specification for formats, sample rate, bit depth, channels, layout, duration, start time, loudness, peaks, naming, metadata, and transfer.
  2. Freeze the approved mix version and source session.
  3. Export full mix, clean, music and effects, dialogue, voice, ambience, or other required stems from identical timing.
  4. Preserve channel order, phase, fades, handles, sync marks, and silence policy.
  5. Measure loudness, true peak, duration, head and tail, channel assignment, and file integrity.
  6. Populate program, episode, language, version, rights, contact, and checksum metadata as required.
  7. Listen to every full mix and spot-check every stem at start, transitions, loud points, and end.
  8. Package files with a manifest, spec, known limitations, and secure delivery route.
  9. Verify transfer hashes and recipient access.

Failure plan

  • If stems do not recombine to the approved mix, correct routing or automation before delivery.
  • If specification is ambiguous, obtain written clarification rather than choosing a destructive default.