Video summary

Audio Post 201: Pro Tools AAF Opening / Dialogue Edit

Main summary

Key takeaways

Educational

Main ideas / concepts covered

  • Live, Q&A-driven workflow for dialogue editing in Avid Pro Tools, using an AAF + video + production audio.
  • Why the session is set up for:
    • Editing first (to make the edit seamless)
    • Processing later (clip gain, alignment, and noise/tick/popup cleanup)
  • Track organization strategy to keep production audio (and VO) manageable while cutting and crossfading shots.
  • Checkerboarding (splitting wide/coverage/shots across separate dialogue tracks) so you can edit by shot coverage and keep mixes consistent.
  • Fill + fade workflow:
    • Smooth playback by ensuring transitions between shots are not audible as hard cuts.
    • Generate ambience (“fill”) when handles are limited.
    • Keep noise reduction/cleanup separate from early “make it seamless” steps.
  • Clip-based processing with iZotope/RX:
    • De-crackle, de-hum, spectral repair (especially for “box”/transient problems).
    • Use RX Monitor / RX Connect for monitoring while editing.
  • Efficient editing choices:
    • Work “top-down” by shot type.
    • Copy/paste fill snippets to reduce repetitive searching.
    • Use extract (unprocessed/rollback-friendly layers) as a safety net before heavy processing.
  • Real-world constraints:
    • Limited time, occasional Pro Tools instability, and balancing speed vs. correctness.

Methodology / step-by-step instructions (detailed)

A) Before editing: stream setup + project acquisition

  • Use OBS to broadcast with adjustable window layouts and better control than YouTube’s live settings.
  • Confirm audio/video in chat (members and viewers indicate if they can hear/see).
  • Explain how the practice project can be downloaded:
    • Website: 4-D digital.com (shop page, practice project)
    • Members may get a coupon (discount).

B) Import and prepare the Pro Tools session

  1. Session creation

    • Create a blank session (not a scoring/template session).
    • Use dialog edit naming and an appropriate start time (e.g., around 59:50).

  2. Import AAF

    • Import the AAF and choose the correct options:
      • Force to session target format
      • Import clip gain
    • Wait for processing and verify audio populates.

  3. Import video

    • Import an encoded DNXHD video version to reduce file size.
    • Recommended conversion tool: Shutter Encoder (H.264 → DNXHD style workflow).
    • Confirm timecode burn-in is visible.

  4. Create audio routing/tracks for monitoring

    • Add new stereo tracks (as needed) for playback/monitoring.
    • Route to a two-bus / output so dialogue can be heard through the stream.

C) Organize OMF/AAF items to reduce complexity

  1. Keep OMF/AAF tracks mostly untouched (reference layer).
  2. Create a dedicated folder structure for dialogue edit:
    • Create folders such as OMF (reference) and Dialogue tracks.
  3. Rename tracks for clarity:
    • Example:
      • Voice over tracks labeled (e.g., “VO”)
      • OMF tracks labeled consistently (e.g., “OMF DX1…DXn”)
  4. Enable clip color coding to match track colors:
    • Preference: default clip coding = track color (avoids messy retagging between sessions).

D) Dialogue edit track plan (checkerboarding philosophy)

  1. Determine number of dialogue tracks needed:
    • Example used: 12 mono dialogue tracks (boom + lavs).
  2. Assign outputs so the mixer can hear them:
    • Route dialogue tracks to the TUX / two-bus (speaker routing noted as key for auditioning).
  3. Create and place VO on its own track(s):
    • Copy from the OMF layer, then VO-muted/unmuted as needed.
  4. Checkerboard approach (conceptual):
    • Break a scene into shot coverage categories (e.g., wide/master, medium closeup, inserts).
    • Place different shots/angles onto different track sets so you can:
      • Keep shots visually consistent
      • Selectively mute/use mics
      • Handle transitions when ambience changes

E) Align / clip gain workflow order (important)

Core principle emphasized:

  • Do fill + edit smoothness first, then align and clip gain afterward (especially for lavs).

Rationale:

  • If you clip gain before alignment, phase/cancellation changes can alter results after AutoAlign.

F) Making transitions seamless (fill + fade workflow)

  1. Clip gain / set baseline for boom first

    • Clip gain boom to a target so the scene’s foundation is consistent.
    • Approximate integrated level targets mentioned:
      • ~ -31 first
      • Later around ~ -27 for scene-level balance

  2. Smooth playback

    • Before noise reduction or heavy plugins:
      • Find where scenes cut (hard/soft cuts)
      • Add short fades (e.g., one-frame fades) at boundaries

  3. Detect missing handles / lack of matching ambience

    • If there isn’t enough naturally matching room tone between shots:
      • Generate ambience fill using tools like Absentia DX

  4. Create “fill” regions

    • Use different sections of room tone (not just one constant segment).
    • Apply fades into/out of fill regions so transitions are not audible.

  5. Copy fill to other shots/tracks where appropriate

    • Reuse transitions to save time when shot locations are acoustically similar.

G) Safety net: use “extracts” / unprocessed reference layers

  • After you have a smooth raw edit:
    • Duplicate or move your assembled edit to extract/XT tracks or inactive versions.
  • Purpose:
    • If heavy processing (AI/noise reduction) makes things worse, you can revert/compare quickly.

H) Auto-Align Post + clip gain for lavs

  • Align lavs to boom references per shot group.
    • Boom references often on tracks like dialog 1/3/5/8 (example structure).
    • Lav tracks aligned to the nearest corresponding boom/dialog track.
  • Then clip gain lavs to match scene loudness (using integrated targets as guidance).

I) Automated “fast cleanup” pass with RX/iZotope (batch when time is short)

  • When needing speed, the workflow shifts to “efficient but not perfect”:
    • Use RX Connect / RX Monitor and iZotope module chains.
  • Emphasis order:
    1. De-crackle
    2. Then dehum
    3. Then targeted spectral repair / declick when needed
  • Mentioned limit/behavior:
    • RX Connect can only process a limited number of clips at once (32 clips mentioned).

J) Spectral repair / declick logic for specific problems

  • For transient issues (e.g., “box handle flip around,” ticks/pops, lighter clicks):
    • Try spectral repair modes:
      • Attenuate / replace variants
    • If spectral repair fails or introduces artifacts:
      • Adjust sensitivity and/or do copy/paste fill from nearby regions
  • For harmonic-heavy objects (silverware, earrings, certain “lighter” clicks):
    • Use a Deconstruct type approach:
      • reduce tonal/transient selectively
      • preserve the noise bed if appropriate
  • For small clicks/harmonics:
    • Use DeClick targeting high-frequency emphasis

K) Production effects vs. dialogue tracks (routing guidance)

  • Distinction:
    • Dialogue tracks: keep essential ambience, steps, breaths where appropriate
    • Production effects (PFX) tracks: split large transient events (e.g., impacts, big hits) for different compression/mixing treatment
  • On bigger productions:
    • Supervisors may leave more production on dialogue tracks for consistency, but the split depends on the team.

L) VO and wild line handling

  • VO is treated as usable/visual-first rather than perfect dialogue fidelity (heavy cleanup is downplayed early for VO).
  • “Wild lines” (production/ADR type pickups):
    • Kept until needed
    • Later used for ADR-style coverage markers (effort markers referenced as “Kyle/Josh effort markers,” etc.)

M) ADR / efforts marker approach (late-stage)

  • Create markers/region groups:
    • e.g., Kyle ADR, Josh ADR
  • Strategy:
    • Identify “who speaks/creates the effort” and where in time it happens.
    • Align efforts like production sound:
      • use fill + fades around action hits
  • Risk management:
    • Pro Tools stability can fail under load; backups are used and the workflow may require restart/force-quit.

Speakers / sources featured

Main speaker

  • The YouTube video host/instructor (audio editor; referenced in the stream, runs the live stream, demonstrates the Pro Tools workflow).

Viewer participants (named during the stream)

  • Kevin
  • Angela
  • Josh (character name within the project)
  • Kyle (character name within the project)
  • Tom Carmen / Tom (viewer question)
  • Angela (viewer question)
  • Robin
  • AR Bob (viewer question)
  • Dell
  • Paul
  • NeuroM Moto
  • ivvan (viewer question/greeting)
  • DTW Productions (viewer question)
  • Mr Shep (viewer question)
  • Dell (again) (additional questions)
  • Saba
  • Hyen
  • Colia / Colombia (greeting/correction)
  • D (greeting/interaction)
  • Ringo
  • macki Mora sound edor (viewer praise)
  • Dell (additional remarks)

Software / systems referenced as tools

  • Pro Tools (Avid) (primary DAW)
  • OBS (streaming software)
  • Shutter Encoder (video transcoding)
  • RX (iZotope/Neutron RX) and RX Connect / RX Monitor (de-noise/cleanup workflows)
  • iZotope modules (e.g., de-crackle/de-hum/declick/deconstruct-style processing)
  • Absentia DX (ambience/fill generation)
  • Auto-Align Post
  • Isotope EQ Match (tone matching of camera audio vs lav)

Project/website source

  • 4-D digital.com (host’s website for practice project download)

Original video