When did you come up with the idea of the Plangent Process?
In the early '80s I was working in New York at Howard Schwartz Recording. Howie wanted to prove that Scotch 250 was the best tape, because it's what he was using; everybody else was using [Ampex] 456. He bought one of the first Hewlett Packard digital frequency synthesizers, which put out a perfect sine wave without drift. He also purchased a very early General Radio spectrum analyzer. I was assigned to sit there and prove that 250 was the best tape. I, being an idiot and not understanding business and politics, proved to him that actually AGFA 468 was the best tape,which wasn't what he had in mind. I learned a lot from the experience! [laughing] But what started this all back then was that I was laying down a 10 kHz tone and then looking at it off tape, seeing how much level the tapes could handle and so on. I noticed that the digital frequency synthesizer would put down a perfect spike in input mode, but in repro off the tape it looked like a pine tree! I asked the tech guys,"What is that? "They said, "That's the wow and flutter." They showed me that if you've got sidebands of 60 Hz on either side of 10 kHz, then you've got a 60 Hz flutter. I was scrubbing the tape very slowly at various points, and I'd hear this whistling noise. I asked, "What is that?" They said, "That's the bias." Because at 30 ips, a 100 kHz bias tone, if you scrub it slow enough, you'll hear it. Many of us have heard this when the tape comes up to speed. I was hearing the bias and that it had a pitch. Meanwhile, I was seeing that the machine was much more messy in the time domain than the digital frequency synthesizer was. I questioned, "Why hasn't anybody used the bias to servo-lock the machine?" Older tapes were recorded with a bias oscillator that was way more stable than the transport was. When you get into the quartz oscillators that you see in the MCI, Studers and later machines, it's pretty awesome how stable they were but how much worse the transport was. I don't think that people realize how much residual wow and flutter we are sensitive to. We're just used to it.
And now you can take that bias and virtually "lock" the audio, post tape.
Yup. It's not easy to do, and the bias tone has a very low signal-to-noise ratio. I took what I'd learned about FM theory from David Smith, who at the time was chief tech for Phil Ramone. David had gotten his Masters in FM theory from RPI [Rensselaer Polytechnic Institute]. Then I went and talked to all the guys that have been in the tape business for many, many years — Dale Manquen [3M/Flying Faders/P&G], John French [JRF Magnetic Sciences], Greg Orton [Ampex, Flux Magnetics], Mike Spitz [Ampex/ATR Services] — and I said to these guys, "What would you have done in 1975 when digital was starting to come in? What if you would have had the opportunity between 1975 and today to improve on the tape playback system? What would the bean-counters not let you build?" So they taught me as much as I could comprehend, and a lot I couldn't. These guys are the analog ninjas! [After that research] I hired a team of great prototypers, and we went ahead and built it with a basic electronic design from David. When they first assembled it, we were very excited. We were all together in a hotel conference room. I had an [Ampex] ATR-102 that had been rebuilt by Mike Spitz,so I knew that it was stable, and I had a tape that was made by MRL that had a 70 kHz bias on it. I stuck a pencil in the machine to make the tape speed go all over the place. My buddy, Dr. Patrick Wolfe, who taught at Cambridge University, UK, [he wrote the math and coded the software] loaded the audio transfer in his laptop. About a minute and a half later, he had ten seconds for us to listen to and it was perfect. It worked on the first try!
So the software had been written?
I had the software written simultaneously to the hardware being manufactured — I took a lot of risks. I thought this would work, so I spent a bunch of bucks on having the preamp made and the software developed. I went out on a limb. Flux Magnetics made the heads; Greg wasn't sure whether they'd work and I told him to go ahead and build one and bill it anyway.
Is there a separate head that picks up the bias?
No, it's all one head. It's a custom single head with a very wideband quiet preamp — built by engineer John Chester — and it's a great playback system.
Does it use a crossover and record on to two separate audio and bias tone tracks?
Yeah, we filter it out separately. Dan Lavry [Lavry Engineering] consulted on the complicated filters. The audio is minimal and transparent. We did everything that we possibly could to get the best possible sound. This is not me — this is the A-team. The DSP [digital signal processing] was written very carefully. It's non real-time. This is not something that you can do with a pitch-shifter, and it's also not something that you can do by analyzing the musical material. You don't know what's vibrato or not. We're working to hundredths of a percent, at 5,000 or 6,000 parts per second, and we can knock out everything the machines did. We see 120, 220, and 440 Hz artifacts all the time — people don't realize how much that's actually contributing to the sound. You also get "scrape flutter." The metal-to-metal contact on the head creates a violin bow effect with the open parts of the tape; the tape "sings" at 3 kHz. And that's modulating all of the music, all the time. Generation loss now appears quite clearly to have been a result of the transport mechanism more than any other thing.
Really?
Yeah, the fast flutter clouds everything. Look, everybody loves the analog sound, especially us, but nobody wants the wow and flutter part of it. The thing that's really been the hardest sell is that obviously you can use Plangent Processes for triage, for stuff that's stretched or where the machine is in trouble; but listen to what it does on a well-made, hi-fi recording from the '60s or the '70s, in terms of clearing out all the debris that was caused by the transport. That was the surprise. Listen to how much better and fresher the imaging is, how much more depth there is, and how solid it feels.One of the more famous producers in the world, known for his EMT plates and chambers, listened to our work and said, "My reverb tails are longer, and that's what it sounded like in the control room. I've never heard that off tape." Something that I want to be really clear about is that we're not "fixing" their work. There have unfortunately been some misapprehensions about our claims with that.We didn't mix the damn record! They mixed the record, and it sounds awesome, but the mixdown machine hurt them a little bit. If we can get that hurt out of there, it's going to sound closer to the console and more like what the original intent was. If Elliot Scheiner, Jimmy Iovine, or any of these wonderful producers are sitting at the board and hearing something and the machine is changing it slightly, we can remove that filter; that's for the better.
Absolutely.
We want to hear what the mixer heard, what he did, and what he tried like hell to get to us when the best transfer mechanism that was possible at the time was tape.
