← All articles
technologysoundfilmrecordingSeptember 17, 20264 min read

What Is Foley Sound? Footsteps Recorded by Someone Watching the Film

By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.

The footsteps, rustling clothes, creaking leather and clinking glasses in a film are almost never the sounds recorded on set. They are performed afterwards by a specialist watching the picture and walking, handling props and moving cloth in a studio, because the location recording captured dialogue and everything else badly or not at all.

Why it is necessary

Location sound recording is optimised for dialogue, using directional microphones placed close to the actors, which means everything else is either too quiet, contaminated by traffic and crew noise, or inconsistent between takes. Dialogue is frequently replaced afterwards as well, which removes the accompanying sound entirely. Films dubbed into other languages need a complete soundtrack with the dialogue stripped out, which is called the music and effects track and is contractually required for international distribution, so every sound other than speech must exist as a separate element that can be retained when the voices are replaced. Beyond necessity, performed sound is simply better: a real footstep is quiet, dull and frequently unremarkable, while a performed one can be given weight, character and emotional colour appropriate to the moment, and audiences accept the enhanced version as more real than the actual recording.

How it is done

The practice is named after Jack Foley, who developed it at Universal from the late 1920s as sound arrived, and the method has changed remarkably little:

  • A Foley stage contains pits of different surfaces, including gravel, sand, concrete, wood, marble, water and leaves, so a performer can walk on whatever the scene requires
  • The artist watches the picture and performs in real time, matching every step and movement as an actor would, which is why they are called performers rather than technicians
  • Footsteps are the largest category, with a different pair of shoes for each character and adjustments for weight, mood, urgency and surface
  • Moves, meaning the rustle of clothing as a body shifts, which is recorded separately and is what makes a character feel physically present
  • Specifics, meaning props and objects: doors, keys, cups, weapons, paper and anything handled on screen
  • Substitution is routine and inventive, with celery snapped for breaking bones, coconut shells long used for hooves, cornflour in a leather pouch for footsteps in snow and a wet chamois for various unpleasant biological sounds

Where it sits in the soundtrack

A finished film soundtrack is built from several distinct layers, each produced by different people. Dialogue, whether recorded on set or replaced later. Foley, covering everything the characters physically do. Hard effects, meaning designed sounds for specific events including explosions, vehicles and anything not producible by a person in a studio, which sound designers construct from libraries and field recordings. Atmospheres or ambiences, the continuous background of a place, which establish where the scene is without anyone noticing. Music. All of these are then balanced in the final mix, where the decisive creative work happens, since the same material can be mixed to feel intimate or overwhelming and the mix decides what the audience attends to at each moment. The convention that a sound must be present whenever an object moves is deliberately broken for effect, and removing all Foley from a scene is a recognised technique for producing unease.

Why it has not been automated

Sound effect libraries containing hundreds of thousands of recordings have existed for decades, and Foley persists because timing and character are the whole point. A library footstep can be placed against a picture and will not have the specific weight, hesitation and rhythm of that character in that moment, and an audience notices the mismatch without being able to identify it. Synchronisation to within a few frames is required before a sound reads as belonging to the action, which is a demanding target when a character crosses a room. Performers also bring interpretation, since a nervous character's movements sound different from a confident one's, and directors give notes about emotional quality rather than about volume. Automated approaches using machine learning to generate sound from video have been demonstrated and produce plausible generic results, which is a different thing from the performed specificity the craft delivers. The economics matter too, since Foley is one of the cheaper stages of post-production and replacing it saves little while risking a great deal.

The takeaway

Location recording captures dialogue and little else usably, and international versions require a complete soundtrack without speech, so footsteps, clothing movement and handled props are performed afterwards by an artist watching the picture on a stage of different surfaces. The practice is named after Jack Foley and dates from the arrival of sound. Substitutions including snapped celery and coconut shells are routine, and the craft persists over sound libraries because timing and emotional character cannot be matched from stock.

Practise this

The Technology track

How our gadgets, computers, code and the internet really work - from your first tap on a screen to programming, AI and cybersecurity, one tiny step at a time.

18 units and 2,148 questions, each with a written explanation. Every unit page shows what it covers and real example questions before you start.