For the wedding scene, I wanted the sound to also be real, and I was particularly taken with the sound of the playback music. So this is when the orchestra that is at the wedding is playing the music, what is actually happening is we are playing back pre-recorded music and the musicians are imitating this music. So that music is not actually being performed at the time, but it's being broadcast to the set through speakers, and it has this wonderful... The sound, the way weddings really sound. And again, this was early days, this was 1971-72, and I then got hold of the actual original recordings of that music, and did a kind of forensic, plastic surgery of blending the captured playback sound, and syncing it up with the actual sound. So that when we went into the mix for the film, I had two soundtracks running simultaneously. One of them, the chaotic sound from the set, and the other one, the real sound of the studio recording. And that meant that we could blend them at the appropriate level to get enough feeling of reality, and yet enough substance from the original recording so that the music didn't sound thin and too artificial. And that's a technique that I had already used to a certain extent in THX 1138, and it was a technique that was really going to come into its own in American Graffiti. My word for this is 'worldizing', because it's taking a sound and exposing it to the acoustics of the world that is being shown by the image. What is this world? Well, it could be the wedding in the garden of a big house, or it could be the gymnasium where a band is playing back, or it could be, you know, anything. Wherever the environment seems to indicate.

The advantage of this is not only does it sound realistic, but you can fine-tune the acoustics to match, and put the dialogue that is running simultaneously at the right level of understandability. It's the sonic equivalent of depth of field in photography, whereas if you're taking a portrait of somebody, generally what we do is use a fairly long focal-length lens, make the person's face be in focus, and then at the right f-stop, we throw the background out of focus, so that as soon as you see the picture, your eye knows exactly what it's supposed to be looking at. If everything is in focus, there's a hesitation where you have to see the person and see the background, and then intellectually make a distinction between the background and the foreground. If you're using depth of field, you do this automatically without thinking about it. And the discovery and use of worldizing is the acoustic equivalent of this. So it had both an aesthetic and a practical function to it. If we had played the music without this acoustic blossoming around it, it would have been harder to understand the dialogue that was being spoken, or we would have had to play the music so low that it would now sound artificial. So this allows you to play the music louder than you would normally play it, and yet because it has this acoustic out-of-focusness around it, you can still hear the dialogue which is clear and sharp, and in focus.