音乐闪避Music Ducking

难度LEVELL2 · 进阶L2 · Intermediate 领域TOPIC音频

先学：Prerequisites: 时间线

学习目标LEARNING OBJECTIVES

学完你能用关键帧在对话出现时压低背景音乐、在间隙抬回来，让对白始终听得清。

概念说明CONCEPTS

音乐闪避（Music Ducking）是有人说话时降低背景音乐音量的混音技术——对话永远优先，音乐负责填空隙，不抢戏。

  • 对话为王：混音里人声（dialogue）最重要，音乐和音效都要为它让路。
  • 手动闪避靠关键帧：在音乐片段的增益线上打关键帧，说话时压下去、间隙抬回来（见 片段增益线）。
  • 跟着波形走：关键帧位置看对话片段的波形起止，而不是死对片段切点——人停顿时往往还有气口。
  • 幅度凭耳朵：说话时音乐可压到 -22 dB 上下，间隙抬回 -12 到 -15 dB 左右，以每句词都听清为准。

操作步骤STEPS

  1. 取消音乐轨的 Mute，回放全片听对话与音乐的平衡
  2. 拖音乐片段的增益线，把整体电平降到对话时合适的位置（约 -22 dB）把音乐整体电平降下来
  3. 在第一段对话开始前，Option / Alt 点击音乐增益线打两个关键帧
  4. 把对话前的线段拖回较高电平（约 -12 dB），按 / 键预览进入对话时的闪避效果对话前后的关键帧与电平差
  5. 在每段对话起止处重复打关键帧，间隙抬到约 -15 dB，直到走完全片

练习PRACTICE

  1. 给一段带旁白的片子做手动闪避：旁白段音乐压到约 -22 dB，两处间隙各抬回约 -15 dB。
  2. 回放全片，专挑"词听不清"的位置微调关键帧，直到每句都清楚。

常见问题FAQ

音乐和人声打架怎么办？

先确认人声电平已统一（见 音频标准化），再把音乐整体压低做闪避——不要反过来一味推高人声。

闪避的关键帧打在哪最合适？

看对话片段的波形：波形起来前一点点降，波形结束后一点点升；对着波形打比对着切点打更自然。

能不能自动闪避？

可以，Fairlight 页面有基于侧链（sidechain）的自动闪避。本阶段先练手动关键帧——最快也最可控，进 Fairlight 再学自动化。

Learning Objectives学习目标

After this card you can use keyframes to lower background music under dialogue and bring it back up in the gaps, so every word stays clear.

Concepts概念说明

Music Ducking is the mixing technique of lowering background music whenever someone speaks — dialogue always comes first; music fills the gaps and never steals the scene.

  • Dialogue is king: voices matter most in a mix; music and effects both make room for them.
  • Manual ducking uses keyframes: add keyframes on the music clip's gain line — down during speech, back up in the gaps (see Clip Gain Line).
  • Follow the waveform: place keyframes where the dialogue waveform starts and stops, not blindly on edit points — pauses often still carry breaths.
  • Amount by ear: music can sit around -22 dB under speech and rise to about -12 to -15 dB in gaps; every line must stay intelligible.

Steps操作步骤

  1. Unmute the music track and play the timeline to hear the dialogue-versus-music balance
  2. Drag the music clip's gain line down to a level that works under dialogue (about -22 dB)Bringing the music's overall level down
  3. Before the first line of dialogue, Option / Alt-click the music gain line to add two keyframes
  4. Drag the segment before them back up (about -12 dB) and press / to preview the duck as speech startsKeyframes and level change around dialogue
  5. Repeat keyframes around every dialogue start and end, raising gaps to about -15 dB, through the whole piece

Practice练习

  1. Manually duck a piece with voiceover: music down to about -22 dB under speech, up to about -15 dB in two gaps.
  2. Play the full piece, hunt for any "can't make out the words" spots, and nudge keyframes until every line is clear.

FAQ常见问题

What do I do when music fights the vocals?

First make sure dialogue levels are consistent (see Normalizing Audio Levels), then lower the music overall and duck it — don't just keep pushing voices louder.

Where exactly should ducking keyframes go?

Watch the dialogue waveform: dip just before it rises, recover just after it ends. Keyframing to the waveform sounds more natural than keyframing to edit points.

Can ducking be automatic?

Yes — the Fairlight page offers sidechain-based auto ducking. For now, practice manual keyframes: fastest and most controllable. Automation comes when you move into Fairlight.

深度资料：音乐闪避 →Deep dive: 音乐闪避 →

本卡术语Terms in this card

时间线Timeline 关键帧Keyframe 配音Voiceover Gain（亮部）Gain 静帧Still Fairlight 音频Fairlight 音效Sound Effects