Sound Effects, Ambience, Music Editing and Video Loudness.

Once dialogue is assembled, the piece still sounds like a sketch: voices over silence, music entering and leaving without reason, and no sound explaining where we are. This stage adds missing layers—effects, ambience and music—and finalises delivery level so the platform need not guess.

This chapter covers Sound Effects, Ambience, Music Editing and Video Loudness: building each layer, cutting music to picture and measuring the final level.

Sound Effects: Where Each Sound Comes From

Sound effects are everything outside dialogue and music: impacts, scrapes, mechanisms, background footsteps, transitions and design details. In practice, three families group them: hard effects occurring on screen, supporting textures, and design effects absent from the real world but giving the piece identity.

A solid effect is usually a composition rather than one file. Combine a transient component marking the hit, a body giving weight and a tail integrating it into space. Check phase before stacking layers: two out-of-phase versions of the same hit consume bass and sound thin despite what the meter says.

Origin is immaterial. Libraries offer variety and professional recordings; recording the shot’s particular sounds matches what you see exactly. Organisation determines working speed: consistent names with source, place and date, and metadata locating a sound without opening twenty. The same discipline used when creating field-recording libraries applies here.

Choose by perspective rather than taste. A close-recorded effect sounds present and detailed; one captured farther away sounds smaller and surrounded by space. A wide shot needs its distance and tail; a close-up needs detail. Combining library effects with different styles without matching distance and reverb gives the typical symptom of sounds seeming to come from different places within one scene.

Ambience: The Layer You Notice When Missing

Ambience is the environmental sound bed sustaining a scene: room air, distant traffic, appliances, wind or venue murmur. Build it in layers—a continuous bed, occasional distant events and close details appearing only when the shot justifies them—so it has depth instead of sounding looped.

Its main function is hiding cuts. When dialogue is edited take by take, the environmental bed maintains continuity and makes joins imperceptible. It therefore follows dialogue and moves with it: when a scene changes day or location, ambience changes on the same frame.

Two recurring mistakes: silence between scenes, more obvious than any noise, and overly prominent ambience competing with voice. The practical rule is listening to ambience with the mouth closed and the voice off-screen: if ambience is only noticed when dialogue stops, it is in place.

Music Editing: Music to Picture

Music editing adjusts existing music to visual structure: cutting, extending and positioning rather than composing. Start with the picture imported into the session and markers at points of interest—scene changes, dramatic hits and logo entrances—where music can enter, leave or change.

The most common editing technique uses one working track for cuts and two alternating music tracks: try each edit on the working track, then move it to either music track so crossfades remain possible without undoing the previous cut. Place cuts on hits or visual transitions; cuts mid-phrase are masked by an effect, short fade or sound-effect entrance.

Also decide music origin. Commercial songs require a licence for any published piece; royalty-free music libraries cover use within their terms. If the picture arrived with temporary music, that reference explains the requested style but is not for delivery.

One professional detail speeds work considerably: prepare the session at 48 kHz with picture at its actual speed, and measure durations in bars before cutting. Cutting solely to frames without following bars makes every adjustment require repositioning; cutting on bars and markers makes moving a whole block one mouse drag. Document what was cut and why: when a revision makes the scene three seconds longer, the note saves reconstructing the reasoning.

Video Loudness: Measure Before Exporting

Video loudness is a destination specification rather than personal taste. Measure it in LUFS or LKFS, integrated over the complete piece alongside a true-peak ceiling; each destination has its target. European television works at -23 LUFS under EBU R 128 and US television at -24 LKFS under ATSC A/85.

Video platforms normalise independently, lowering or raising the mix to their target, around -14 LUFS for YouTube. That does not mean delivering at -14; understand that an excessively quiet mix will be raised, exposing background noise, while an excessively loud mix will be clipped.

Measuring before export resolves both: review the whole mix’s integrated loudness, true peak with safety margin and consistency between scenes, then correct using level automation rather than an overworked limiter. Metering and headroom details are in mastering, true peak and limiting, and the same criterion applies to radio and podcasting.

Balance the Mix Layers

A picture mix has a simple hierarchy: voice first, its support second, decoration last. Move music and effects around dialogue through automation—down during important phrases, up in pauses—rather than hoping a compressor does that work blindly.

Three checks avoid most problems. Listen at a constant reference level, because hearing adapts and everything seems correct after five minutes. Check in mono for elements existing only in one channel and disappearing on single-speaker displays. Compare with a reference of similar volume and style, rather than judging against the memory of the previous mix.

Picture also directs the stereo field. If a sound is anchored to an object crossing the shot, pan with it; ambience occupies full width without sudden movement; something accompanying voice stays central to avoid competition. Unmotivated stereo movement distracts from picture, and is noticed much sooner on small channels or phones than in the studio.

Video Audio Delivery Checklist

Before exporting: sample rate matching the video project, sufficient bit depth, no clipped peaks, integrated loudness within the destination target, true peak with headroom, the first and last half-second checked, and synchronisation verified against picture from beginning to end.

Export the master in the destination’s required format, retain the project and individual elements for corrections, and note the delivered level. Recording the level seems excessive until the first revision asks what level the mix used.

Next Step

With the complete soundtrack and measured level, finalise synchronisation and formal delivery: timecode, stems, channel specifications and delivery packages. These are covered in the guide to audio/video sync, timecode, deliverables, stems and broadcast specs.