Sound design

#Moving layers

Moving layers lists the layers in the scene that move, with a count at the top. Each row has chips that say why it is there: Particles, Displace (Motion's per-particle move), Gradient, Pose (it changes across linked States) or Layout (Motion In or Motion Out). If nothing moves yet, the panel says so and offers Open Canvas.

Today only a Particles row can make sound: a Vector layer with particles switched on and moving. The other rows show so you can see what moves, but their own voice (Motion Voice) is still planned, and their card says so. A Vector without particles tells you to add Count and then Motion in Canvas to give it a voice. See Particles for that.

Click a row to open its card. The switch on the row turns the layer's voice on or off across all its linked States. A layer with its voice on always stays in the list, even if it stops moving, so you can switch it off.

Growing a layer's sound

A layer's sound is a chain you build with two buttons:

• + Add Particle Voice adds a voice under the last one. Choose Synth (movement-first whooshes or an instrument) or Granular (grains of a sample you generate or upload). A layer can have up to 12 voices.
• + Add Component adds an effect for the whole layer: Resonate, Zones, Reverb and, once there are two or more voices, FM. Each can be added once.

Close a card with its x; an Undo toast lets you bring it back, with its settings.

#Particle Voice

A Particle Voice follows one particle. Its speed sets the level, brightness and vibrato, its height tints the pitch, and its position from left to right sets the pan. A new voice starts as a whoosh on the particle farthest from the layer's other voices, so two voices never sit on the same spot. If the particle engine is off, the voice waits silently until the particles move.

The voice card

• Compose rolls a new Seed and, with it, a new tone. Type an old seed and press Enter to get that exact tone back.
• Play auditions the voice.
• Index picks the particle to follow; its ring glows on the ad. Highlight sets how bright that glow is. The glow is only in the editor, never in an export.
• Chaos gives the voice a life of its own. At 0 it sounds from the first frame to the last. As it rises, the voice triggers on movement, swells, decays, rests, and now and then hops to another particle. It is seeded, so the export sounds like the preview; the dice button rolls a different performance at the same Chaos.
• Speed to vol sets how strongly speed drives loudness, and Response how tightly the sound follows the motion.
• Grid overlay shows the note bands on the ad so you can see which height plays which pitch.
• Solo lets you hear only the soloed voices while you edit (it never changes an export). Mute silences a voice in the preview and in exports.

Synth

Pick a Voice: Celeste, Warm pad, Vocal aah, Bell, Pluck, Saw, Organ, Glass, Noise/whoosh or Air. Scale, Key and Octave set the pitch tint; if a whoosh feels thin, drop the octave. The shape sliders tune it: Body (the resting tone), Whoosh (how far it opens on a fast pass), Q (how ringy), Air (less rumble), Pitch tint, plus Brightness, Vibrato and Glide. Reset shape puts them back.

Granular

Describe a sound ("glassy metallic ping") and press Generate sound, or Upload your own file. Hold Hold to hear the sample. The particle then plays it as grains: left to right scrubs through the sample and height sets the pitch. Grain size, Density and Spray go from fine texture to a smooth wash.

Layer effects

• Resonate is a tuned, metallic delay: moving side to side sweeps the delay, height sets its pitch. Feedback sets how long the echoes ring.
• Zones splits the space around the ad into eight zones (edges and corners). Assign a setting to a zone, and particles moving into it push that setting up. Randomize fills the zones from the effects you have added.
• Reverb puts space around every voice on the layer.
• FM uses the last two voices: the last one carries, the one before it modulates, and the faster the modulator's particle moves, the harder it bends. Ratio at 1, 2 or 3.5 stays in tune; 1.41 and 2.76 clang like bells and metal.

Worked example

A 1:1 ad where a particle cloud bursts outward and settles into the product.

1. Add a Synth voice set to Noise/whoosh, with Chaos at 0 so it sounds the whole scene.
2. Add a second voice set to Bell on a particle near the centre, with Chaos raised so it rings only when it moves.
3. Add Reverb, then Zones with the top edge pushing the Bell's brightness.
4. Press Play. The burst whooshes and the Bell sparkles as particles fly past the top.

#Audio Library

The Audio Library keeps the brand's music in one place for every project on the design system. It fills itself: drop an audio file on the preview and it becomes the song for the whole timeline; drop one on a scene's bar in the timeline and it plays under that scene only. Either way it joins the library, once.

In another project on the same design system, press Use on a track. With the timeline wave up (no scene bar focused) it becomes the timeline song, whole, just like a drop on the preview. With a scene's bar focused it goes under that scene only. The panel shows Timeline Audio (with Volume and Remove) when no scene is focused, and Scene Volume with that scene's audio when you focus a scene's bar.

#Full-length songs

A song keeps its full length, however long: a 5 minute track for a music visualizer imports whole, the waveform shows all of it (switch the wave to Fit to zoom in on the highlight), and Play, Render Linked Scenes and Render Timeline all play it continuously across the scenes, never restarting at a cut.

What plays is the part of the song the scenes cover: the highlight on the waveform is as long as the timeline. A scene still lasts 0.1 to 30 seconds, so on a one-scene project the highlight is at most 30 seconds; add scenes and it grows with them until it spans the whole song.

Under the track name, Timeline Audio says where you are, for example "Song 3:42 · scenes 0:30 · 3:12 not covered yet". To fill the rest in one go:

1. Drop the song on the preview (or press Use on it in the Audio Library).
2. In Timeline Audio, set every to the scene length you want (30 s makes the fewest scenes), and press Add scenes to cover the song. The new scenes are copies of the last scene, linked to it, and the last one is trimmed so the timeline ends with the song. One Undo takes them all back.
3. Chop it by hand: drag the scene edges on the timeline to land the cuts where you want them, or split a scene at the playhead by pressing + Add Scene (or Cmd+A) while it plays.

A project holds up to 60 scenes, so with 30 second scenes a song can run to 30 minutes; with 5 second scenes, to 5 minutes. A song dropped on a single scene's bar plays only under that scene; the scene panel then offers Use the whole song on the timeline to move it.

Uploads: a song up to 500 MB goes straight to storage from the browser (a 10 minute WAV is about 100 MB), and an agent's upload_audio imports one of the same size from a URL.

#Captured video audio

When a scene holds a video clip that carries sound, this section lists it (the background video or the layer by name) with a switch that is muted by default. Most ad placements autoplay without sound, and picture-only clips are the safer default, so you turn on only the clips whose sound you mean to keep. Footage you capture with Live Frame has no sound, so it does not appear here.

#Beat sync

Beat sync turns a song into a cut list. Start a project with one scene and a song on the main timeline, then:

1. Drag the highlight window on the waveform over the part of the song you want.
2. Press Detect transients. The readout shows how many hits it found and how many scenes that makes.
3. Wildfield adds scenes to match. Each new scene lasts the length of the window divided by the number of hits, so the cuts follow the song's average beat.

It works on the main-timeline song only, so drop the track on the preview first. It is shown only while the project has a single scene: use it at the start, before you design the scenes.

#Master volume

The speaker button in the timeline bar opens Master Volume: a slider from 0 to 100%, saved with the project. It sets the level of everything together (voices, the timeline song, scene audio and video sound), and it applies to the export as well as the preview. At 0 the icon turns red. To silence one voice, use its Mute instead; to silence a layer, turn off its Voice switch.

Sound reaches the video through Include sound in Export, which bakes the voices into the MP4 so it carries exactly what the preview plays. Muted voices are left out and Solo is ignored.