Preliminary mix of individual tracks before the final mix — reduces track count and establishes an early balance.
Technical Details
A typical predub is divided into three to six main groups: Dialogue/ADR (usually 8-16 tracks), Music (4-8 tracks), Atmospheres/Ambiences (6-12 tracks), Hard Effects/SFX (12-24 tracks), Foley (4-8 tracks), and occasionally a separate group for Low-Frequency Effects (LFE). Each predub is output on separate outputs from the Digital Audio Workstation (Pro Tools, Nuendo) and rendered as 48kHz/24Bit files. The stems retain their original channel configuration – a 5.1 predub thus generates six discrete mono files per group.
History & Development
The predub process was established in the 1940s with the introduction of magnetic sound recording in Hollywood studios. Previously, all sound tracks had to be mixed live in a single pass. MGM and Warner Bros. perfected the three-stage mixing process from 1952 onwards: Predub – Temp Dub – Final Mix. With digitalization from the 1980s, systems like the Fairlight CMI enabled more precise editing of predubs through non-destructive automation and recall functions.
Practical Application in Film
For "Mad Max: Fury Road" (2015), Sound Designer Mark Mangini created over 40 separate predubs solely for vehicle sounds to keep each car individually controllable. Christopher Nolan's films systematically use dialogue predubs with various reverb variants, which are only selected in the final mix depending on the editing pace. In complex productions, this workflow saves up to 30% of mixing time, as corrections only need to be made in the affected predub group without impacting the entire mix.
Comparison & Alternatives
Predubs differ from submixes by their final, render-based nature – a submix remains editable, a predub is fixed. Track layering works with ungrouped individual tracks and requires more processing power in the final mix. Modern object-based audio systems like Dolby Atmos reduce the predub effort, as objects remain individually positionable. Smaller productions often forgo predubs and work directly with submix groups, but lose the ability for precise stem mastering for various distribution formats.