6 min read
What is sound design, and where does it start in a film?
Not just explosions and whooshes. Sound design is the decision about what the audience hears, and it starts long before post.
Sound design is the craft of deciding what an audience hears and building it. On a film that covers effects, ambiences, foley, processed textures, silence and the way all of it sits against dialogue and music. On a game or an installation it also covers how sound reacts to what the user does.
What it includes
- Hard effects: doors, cars, guns, impacts, anything tied to an action on screen.
- Ambiences: the room, the street, the forest, the thing that tells you where you are.
- Foley: footsteps, cloth, props, performed to picture.
- Designed elements: sounds that do not exist and have to be built from layers.
- Spotting and structure: choosing where sound leans in and where it steps back.
What it does not include
Dialogue editing, ADR and the final mix are separate crafts, even when one person does all of them on a small production. Music is usually its own department too. Sound design is the layer between them, and the good version is written to leave room for both.
Where it starts
On paper. A script that says the machine wakes up is a sound design brief. Reading the script before the shoot tells you what to record on location, what wild tracks you need and what will have to be built later. The most expensive sound design is the kind invented in the last week of post.
Why it matters more than the budget suggests
Sound carries emotion faster than picture and costs a fraction of a VFX shot. Audiences forgive a soft image and abandon a film with bad sound. A scene cut with a considered ambience and three well placed effects reads as expensive, whatever the camera was.
The technical floor
Design decisions only survive if the files are clean. Work at 48 kHz, 24 bit, keep everything mono until you have a reason not to, and check the real properties of every source rather than trusting the filename. The audio inspector on this site reads sample rate, bit depth, channel layout and metadata locally, which is the fastest way to catch a 44.1 kHz library file before it drifts in your timeline.