A studio recording is too clean to belong anywhere. Worldizing is the old fix: play it out loud in a real space, record it back, and let the room do the thing no preset quite manages. Sound editors have leaned on it since the seventies.
Where it came from, why it beats a reverb plugin, how to do it digitally, and when it is actually worth the trouble.
Walter Murch coined it on American Graffiti in 1973. The film is wall to wall with pop songs coming out of car radios and garages and drive-ins, and cut against picture the clean record masters sounded exactly like what they were: clean record masters, dropped on top of the scene.
So Murch took them back outside. He played the songs through a speaker in a real neighbourhood and re-recorded them with a microphone some distance away, moving around while the tape rolled. What came back carried the actual air of those places, the reflections off real houses and cars and tarmac. He called it worldizing. You take a sound that lives nowhere and you put it into the world.
A room is not a reverb tail. That is the part the plugin misses.
What your ear actually uses to place a sound is a pile of small, correlated cues arriving together: the first reflections off the near surfaces, the specific way the room soaks up highs, where the source is standing, where you are standing, a little background you barely clock. Record in a real space and every one of those turns up for free, in agreement with each other, exactly the way a lifetime of listening has taught you to expect.
A preset gives you a believable tail and almost none of the rest. That is the whole reason worldized audio drops into a scene and sits there, while a track with a hall on it tends to float an inch above the picture.
You need a speaker, something to record with, a microphone, and a space worth capturing. Then:
| Step | What you do |
|---|---|
| Play it | Send the clean track out of the speaker, into the real room |
| Catch it | Put the mic where the listener would be, and record the room's version |
| Move | Shift the speaker or the mic for distance and angle. Further and off-axis for "down the hall" |
| Layer | Do a close pass and a far pass. Blending the two is what gives you control later |
It sounds fussy because it is. You are running a small recording session, you need the actual room, and once it is on tape the space is baked in. There is no undo. That last point matters more than it looks, and it is where the digital version earns its place.
The modern version swaps the live re-recording for an impulse response. You capture a room once, by playing a sine sweep or popping a balloon and recording the result, then software deconvolves that into the room's acoustic fingerprint. Load the fingerprint into a convolution reverb and now any audio you feed it "plays" in that room.
Same physics, but repeatable, and you can bottle a space you will never get a crew back into. Capture a stairwell once and you own its sound for good.
Audio Ease's Speakerphone is the well known library-driven take on this, built around hundreds of captured speakers and spaces. You do not need it to start though. A convolution reverb and a handful of your own impulse responses cover the core of the idea, and capturing your own is cheap: a measurement mic, a laptop, free software, and a room you like.
Worldizing is for sound that has to belong to a place. Source music leaking from a room, a tannoy in a station, a song on a car stereo, dialogue heard through a wall. Anything the characters can hear, coming from somewhere you could point to. If you are unsure whether a cue counts, that is the line between diegetic and non-diegetic sound, and it is worth getting straight first.
It fights you on score. Non-diegetic music lives outside the world of the film, so putting it in a room works directly against what it is for. Worldize the band at the party. Leave the strings that only the audience can hear well alone.
Play only the worldized sound, no picture. If it sounds like something happening in a place nearby, it has worked. If it sounds like your track with a nice reverb on it, the room is not really there yet, and a real impulse response will do more than a bigger preset.
Whether you did it with a speaker in a street or with an impulse response, you have baked in one room, heard from one position. The second the shot moves toward the sound, that fixed capture stops being true, and you are back to recording another pass or drawing automation.
There is another way to think about it. Instead of capturing one fixed result, describe the geometry, where the sound is and where the listener is, and let the distance, the walls and the doorway between them produce the muffling and the space. Move the listener and it all changes, because the geometry changed and not because you keyframed it.
That is what we built THALM for. You place the music in a room on a building cutaway, put the listening position somewhere else, and drag it. Walk it through a wall and you hear the wall. If you have impulse responses of real rooms, you can drop them straight onto the spaces on the map, so it is your captured worldizing, but with the listener free to move through it.
There is a record mode as well: arm it, press play, drag the music through the building while the scene runs, and it keeps that pass as an editable route locked to the playhead. One take instead of a session with a speaker in the driveway.
It loads as an Audio Unit or VST3, so it drops straight into Final Cut, Premiere, Resolve, Pro Tools and any DAW.
No download and no signup for the browser version. About thirty seconds to know whether it is any use to you.
We write these when there is something worth writing down. One email when a new one lands or a new Tunary instrument ships. No newsletter, no schedule.
Tunary