Issue link: https://digital.copcomm.com/i/1541612
C I N E M A A U D I O S O C I E T Y. O R G 47 Yeah! So, he has some hits, and he's getting a little bit of money and he realizes that this process is destructive. If he messes up, he's got to start all over again. Or if he doesn't like the performance or anything like that, you know? He then graduates to two tape machines and just plays onto one, rewinds it, plays it back to the other. Actually, he starts with two lathes and then goes to two tape machines! Again, he kept everything. As we started going through the archive, we started seeing pieces of all the songs. And nothing was next to each other because he would go, "I need a piece of tape that's a minute long, I'll find it on another reel." So, the collection is like a jigsaw puzzle. Once we realized that those parts were there, we assembled them. Out of the 33,000 recordings, we had about 200 songs in the end with maybe 40 or 50 that were unreleased. He left an amazing amount of material behind. How did you figure out what you had? The initial thought was that if there's 10 or 20 stems to a song—and we call them stems, but they're just mono bounces back-and-forth. If we line them all up and flip the polarity of every other one and try to phase out the thing, we should hear what's new and what's something that's repeated somewhere else. This was a good idea in theory, but the machines drifted so much that it never worked out. You know, there are archivists who undo that exact kind of drift stuff… Well, that's what we did! There's an archivist who works for Harvard's version of the Library of Congress, Bruce Gordon, who does all the audio media capture. He's the first person I talked to about fixing drift when he was doing old transfers and he used some Cedar technology. [Ed. See the Winter 2024 CAS Quarterly for Patrick's interview with Bruce.] Cedar has a process called ReSpeed that we use. It basically reads the bias line and if the bias line dips, it corrects it. The problem with Les's recordings is they could have 30 bias lines because he's going back-and-forth and back-and-forth between machines. The Cedar software would just go nuts, freak out. So, I started working with those guys, and we figured out a way to disable that part and manually do it. Wow! Yeah, and I can tell you it's incredibly time-consuming. Out of the 200, we've done around 100. And it takes about a day for each one. It's been about four- or five-months' worth of work just to get these prepared so that we can start to mix! Here, I'll show you. [Points to a Pro Tools setup and plays the session.] This first clip would be his first step. So, he played bass and then went back and played guitar on top of it. Then, on the next track, you still have the bass and the guitar, but now you'll see that Mary [Ford, his wife, who was a singer and guitarist] starts singing. I'm hearing reverb on her vocal. Did he have a chamber? Yes, this one is in his home in Mahwah, New Jersey. He started his whole process out here in L.A. and then, I believe in the early '50s, would have moved back to New Jersey, which is where Les stayed for the rest of his life. We even went to the house and did impulse responses of the chamber, so that's what I'm using to do the work that we do now. Very cool. In fact, I just had Bill Putnam Jr. here about two days ago, and he's going to go out to New Jersey and do a real thorough IR shooting for us because the house is rotting and it's going to fall down. He wants to do the live room and the chamber and… So, Les had a full-blown studio at home? They weren't just recording in the living room? No, no, I'll show you a picture. This would be exactly how he's doing it. This is the house in New Jersey, it's the live room. You see one tape machine and there'd be another one over here. He's going back-and-forth, just using that mixer. Ampex 300