Working Video & audio skill
Mix a podcast episode
Mix a podcast episode for intelligible dialogue, consistent perspective, controlled noise, licensed elements, and platform-ready loudness.
Protect speech meaning before pursuing polish.
When to use
- Use after editorial structure and source selections are approved.
- Preserve the session and untreated dialogue before destructive repair.
- Do not use repair or editing to change a speaker's meaning, identity, or consent boundary.
Procedure
- Confirm sample rate, channel layout, edit lock, deliverables, loudness target, true-peak limit, and licensed media.
- Organize dialogue, room tone, ambience, music, effects, ads, and pickups with stable labels.
- Repair clicks, hum, plosives, clipping, dropouts, and noise conservatively; compare against the original.
- Match dialogue tone, level, perspective, and room continuity without erasing natural speech.
- Use automation for intelligibility, breaths, overlaps, music ducking, transitions, and emphasis.
- Check phase, mono compatibility, stereo image, sibilance, low end, and listener fatigue.
- Measure integrated and short-term loudness plus true peaks across full program and inserted segments.
- Listen on reference monitors, headphones, phone speaker, and in mono from start to finish.
- Export master and distribution files with metadata, checksum, and QC record.
Failure plan
- If repair changes words or identity, restore the source and choose a transparent edit or pickup.
- If an inserted segment breaks delivery limits, remix the segment rather than limiting the entire episode blindly.
--- name: mix-a-podcast-episode category: media description: Mix a podcast episode for intelligible dialogue, consistent perspective, controlled noise, licensed elements, and platform-ready loudness. Use when edited speech, music, ambience, and effects must become one reliable master. --- # mix-a-podcast-episode Protect speech meaning before pursuing polish. ## When to use - Use after editorial structure and source selections are approved. - Preserve the session and untreated dialogue before destructive repair. - Do not use repair or editing to change a speaker's meaning, identity, or consent boundary. ## Procedure 1. Confirm sample rate, channel layout, edit lock, deliverables, loudness target, true-peak limit, and licensed media. 2. Organize dialogue, room tone, ambience, music, effects, ads, and pickups with stable labels. 3. Repair clicks, hum, plosives, clipping, dropouts, and noise conservatively; compare against the original. 4. Match dialogue tone, level, perspective, and room continuity without erasing natural speech. 5. Use automation for intelligibility, breaths, overlaps, music ducking, transitions, and emphasis. 6. Check phase, mono compatibility, stereo image, sibilance, low end, and listener fatigue. 7. Measure integrated and short-term loudness plus true peaks across full program and inserted segments. 8. Listen on reference monitors, headphones, phone speaker, and in mono from start to finish. 9. Export master and distribution files with metadata, checksum, and QC record. ## Failure plan - If repair changes words or identity, restore the source and choose a transparent edit or pickup. - If an inserted segment breaks delivery limits, remix the segment rather than limiting the entire episode blindly. ## Done - A mix report records session version, sources, repairs, processing, rights, loudness, peaks, formats, metadata, and limitations - Full-length playback and technical checks verify intelligibility, continuity, mono, artifacts, loudness, true peak, and exported files