Skip to content

description: "Premiere Audio Mixing & Sound Design — Level 1 (Module 8) covered per-clip volume and the Essential Sound panel's quick presets. This module goes further…"---

06 · Premiere Audio Mixing & Sound Design

Level 1 (Module 8) covered per-clip volume and the Essential Sound panel's quick presets. This module goes further into multi-track mixing with the Audio Track Mixer, submixes and sends, and basic sound design — layering ambience/SFX and ducking music under dialogue — the pass that makes a video sound as polished as it looks.

1. The Audio Track Mixer

Where the Essential Sound panel works clip-by-clip, the Audio Track Mixer (Window > Audio Track Mixer) gives you one full channel strip per track, mirroring a physical mixing console — the right tool once a sequence has several tracks that need to be balanced against each other as a whole.

  1. Open Window > Audio Track Mixer — it shows one vertical strip per audio track in the active sequence, each with a volume fader, a pan knob, and Mute (M) / Solo (S) / Record-enable (R) buttons matching the equivalent controls on that track's header in the Timeline panel.
  2. Drag a track's fader up/down during playback to adjust its overall level; the same fader also appears (in miniature) on the track header in the Timeline panel itself if the mixer isn't open.
  3. Each strip's automation mode dropdown (top of the strip: Off / Read / Latch / Touch / Write) controls whether moving the fader during playback records a volume change over time as keyframes: Write records continuously while you move it, Touch records only while you're actively dragging and snaps back to the previous level once released, and Read just plays back whatever keyframes already exist without recording new ones.
  4. Solo (S) individual tracks to check dialogue clarity, music, and SFX in isolation before judging how they sit together in the full mix.

2. Submixes and sends

A submix track groups several regular tracks so they can be adjusted, processed, or muted together as one — for example, three separate SFX tracks all routed to one "SFX Submix" so a single fader controls their combined level.

  1. In the Timeline panel, right-click any track header and choose Add Track.... In the dialog, set Track Type to Submix rather than Standard, and choose its channel routing (Stereo is typical for most projects).
  2. On each regular track you want grouped, set its output routing (available in the Audio Track Mixer strip, near the bottom) to point to the new submix instead of the sequence's Master — audio from that track now passes through the submix's fader before reaching Master.
  3. Adjust the submix strip's single fader in the Audio Track Mixer to raise or lower every routed track together, in the same proportion — useful for a fast "all SFX down 3dB" type change without touching several individual track faders one at a time.
  4. A send (also visible per-strip in the Audio Track Mixer) forwards a copy of a track's signal to another track or submix at an independently adjustable level, without removing it from its original track — used less often in basic video sound design than direct output routing, but useful for feeding a shared reverb or effect submix from several tracks at once.

3. Sound design basics

  1. Add ambience/room tone and sound effects as their own clips on dedicated tracks (e.g. A2 for ambience, A3 for SFX), separate from dialogue (A1) — keeping categories on separate tracks is what makes track-level mixing (Sections 1-2) and per-category adjustments practical.
  2. Fade a clip in/out cleanly using the small circular fade handles at its top-left/top-right corners in the timeline (drag inward from the corner) rather than a hard cut, for anything ambient that shouldn't start or stop abruptly.
  3. Open Window > Essential Sound, select a music clip, tag it as Music, and check Duck Against Dialogue — Premiere automatically lowers that music clip's volume wherever dialogue-tagged clips play, and raises it back afterward. Click Generate Keyframes to bake the ducking in as adjustable keyframes on the clip rather than a live, recalculated effect.
  4. For deeper repair (removing background noise, hiss, or a persistent hum) beyond what Essential Sound's Repair section handles, right-click a clip and choose Edit Clip in Adobe Audition (if Audition is installed) — it opens the original audio in Audition for detailed restoration tools, and saving there updates the clip back in Premiere automatically, the same round-trip logic as a Photoshop Smart Object (Level 2, Module 1).
  5. Check final levels with Window > Audio Meters (Level 1, Module 8): dialogue peaking around -12 to -6 dB, music sitting noticeably lower under dialogue (a duck of roughly -15 to -20 dB is a common starting point), and nothing clipping at 0 dB across the full mix.

4. Worked example: dialogue with music and ambience

  1. Place a dialogue clip on A1, a low room-tone/ambience clip on A2 running underneath the whole scene, and a music bed on A3.
  2. Tag the dialogue clip as Dialogue and the music clip as Music in Essential Sound; check Duck Against Dialogue on the music clip and click Generate Keyframes.
  3. Fade the ambience clip in and out at the scene's edges using its corner fade handles so it doesn't start/stop abruptly.
  4. Route A2 and A3 to a shared Ambience/Music Submix track, and use that submix's single fader in the Audio Track Mixer to balance the whole non-dialogue layer against the dialogue track in one move.
  5. Solo A1 to confirm dialogue reads cleanly on its own, then un-solo and check the full mix on Audio Meters for any clipping.

Cheat sheet

Action Where
Audio Track Mixer Window > Audio Track Mixer
Add a submix track Right-click track header > Add Track > Track Type: Submix
Route a track to a submix Audio Track Mixer strip > output routing
Automation mode (record fader moves) Track Mixer strip > mode dropdown (Read/Latch/Touch/Write)
Duck music under dialogue Essential Sound > Music clip > Duck Against Dialogue
Bake ducking into keyframes Essential Sound > Generate Keyframes
Fade a clip in/out Drag the corner fade handles on the clip
Edit audio in Adobe Audition Right-click clip > Edit Clip in Adobe Audition
Check final levels Window > Audio Meters

How It Actually Works

  • The signal flow is a literal summing chain: clip → track → submix → Master, each stage multiplying the signal by a gain value. Every track's fader is a multiplier applied to whatever audio arrives at that track (already summed from every clip on it); routing a track's output to a submix instead of Master means that track's signal is summed together with every other track pointed at the same submix before the submix's own fader multiplies the combined result again on its way to Master. This is exactly why moving one submix fader changes every routed track by the same proportion at once — they've already been mixed into one shared signal by the time that fader touches it, whereas the individual tracks underneath, being still separately routed, remain independently adjustable at their own strips.
  • Automation modes determine whether fader movement is read as data-in or data-out of a keyframe timeline. Write and Touch modes sample the fader's position continuously (or only while touched) and record each sample as a volume keyframe at that point in time — the same underlying keyframe data model from After Effects/Essential Graphics animation. Read mode instead only plays back whatever keyframes already exist, applying their stored values without generating new ones — which is why switching a strip to Read after recording an automation pass gives you a locked, repeatable result on every subsequent playback.
  • Duck Against Dialogue works by automatically generating volume keyframes at every dialogue clip's boundaries, based on tag metadata, not by "listening" to the mix live. Essential Sound scans the sequence for clips tagged Dialogue, and for each one, computes a ramp-down keyframe shortly before that clip starts and a ramp-up keyframe shortly after it ends on the tagged Music clip — Generate Keyframes is the step that actually writes those computed values as literal keyframes onto the music clip's volume property, after which it behaves exactly like any other manually-set volume automation and can be nudged by hand.
  • A fade handle is a shorthand UI for the same volume-keyframe mechanism, constrained to a clip's own local timeline. Dragging a corner fade handle inward writes two keyframes — one at 0% volume at the clip's very edge, one at 100% volume at the point you dragged to — and Premiere interpolates linearly between them, identical in kind to any other keyframe-driven property change, just scoped to that one clip rather than the whole track.
  • dB is a logarithmic ratio, which is why "-15 to -20 dB under dialogue" is a much bigger perceived drop than the numbers suggest. Decibels express a ratio of amplitudes on a logarithmic scale (roughly, every -6 dB halves the signal's amplitude); a -18 dB duck represents the music's amplitude dropping to roughly one-eighth of its unducked level, not a linear "18% quieter" — which is the actual reason a duck that looks like a small fader move produces a dramatic, clearly audible drop under dialogue.

Exercise

Build a short sequence with dialogue on one track, ambience on a second, and music on a third. Tag the clips in Essential Sound, enable Duck Against Dialogue on the music, and generate keyframes. Route the ambience and music tracks to a shared submix and balance them against dialogue using the submix's fader in the Audio Track Mixer. Confirm on Audio Meters that dialogue peaks around -12 to -6 dB with no clipping anywhere in the mix.