by BRUCE BARTLETT WITH JENNY BARTLETT
Hidden away in a small company near Cambridge, England, is a man who upgrades analog tape recorders to near perfection. Tim de Paravicini modifies classic tape machines. He re works them from the ground up. adding new tape heads, tweaking the transport, and replacing the electronics with his own custom tube designs. He also upgrades stereo microphones with special circuitry. And the results are worth it. His customers rave about the beautiful new sound, better than they can get with digital recorders.
An example of the de Paravicini sound is on A Meeting by the River, by Ry Cooder and Vishwa Bhwat. The experience of hearing this CD can be startling because of its 'presence, warmth, and purity. The recording won a Grammy for its engineering by Kavi Alexander of Water Lily Acoustics, an audiophile label. The electronics that contributed to this disc's quality sound are the work of a man with a passion for audio purity that goes beyond digital quality. In fact, de Paravicini has contempt for digital sound as it now exists. He's pushing the envelope far be yond what we ordinarily settle for in CD quality.
If you're fascinated by the search for the ultimate in recording quality, de Paravicini's story of modifications of recorders, consoles, and microphones will interest you, as will his provocative opinions on everything from digital sound to hi-fi "tweaks." A man of many talents, de Paravicini is well known in esoteric audio circles as an extra-ordinary designer. While some of his circuit designs are tube and some transistor, all are original. He has been a consultant to Musical Fidelity and Luc Corporation in the design of many of their classic audio components.
Fourteen years ago, de Paravicini founded Esoteric Audio Research. The company makes vacuum-tube products for the high-end audio and pro-audio markets, such as power amplifiers, a microphone. a microphone amplifier, equalizers, and a compressor/limiter. With years of experience in disc cutting, de Paravicini has cut records for Water Lily Acoustics, Chesky Records, and Island Records. He also improved the performance of several recording lathes.
Although de Paravicini will upgrade any tape machine, his favorite is the legendary Studer C37. "It's a good, reliable workhorse," he says. According to him, the C37 blows away everything else, even an Ampex MR70. The C37's tube circuitry is simple, with no microprocessors to get in the way. Even before the upgrade, the C37's specs are no table. The frequency response is rated as 30 Hz to 15 kHz, +1,-2 dB, and S/N is 75-dB (rms, weighted) at 15 ips. After de Paravicini's modification, the response is 7 Hz to 35 kHz. ±1 dB, and S/N is 90 dB! One happy user of a de Paravicini tape deck is Chris Rice. owner of Altarus Records, a classical label. Rice had de Paravicini modify thee Studer C37s-two 1-inch and one half-inch. "The heads were custom-made to Tim's specifications," says Rice. "The mechanical modifications he did himself. He stripped the electronics out and rebuilt his own circuitry into the existing modules, doing hundreds of modifications. He uses his own EQ curve. He also provides an a.c. mains regenerating power supply because the machines are not quartz locked; they depend on mains frequency. By doubling the capstan diameter, Tim doubled the tape speed from 7 1/2", 15 ips to 15/30 ips." According to Rice, "The new machines are incredibly stable mechanically. They sound fabulous, very quiet and--more dynamic. They're very dean one have no grungy noises. With the new Ampex 499 tape. the. modulation noise is down so low, .you can almost forget about it. It give you quite a lot of leeway in your recording level. You don't have to worry about compression. because these tapes will happily go up to 9 dB over. The tape saturates way before the electronics overload."
Rice notes that his analog decks sound better to him thaw digital, even without any noise reduction. "They give a more accurate representation of what's coming down the line. That has to be my final criterion: How closely can I capture what's coming in from the microphones?"
Another satisfied user is Sam Rivers, a producer of jazz records. He sent de Paravicini two broken-down Studer A-80s to modify into a 1-inch and half-inch model. Vince Clark of Erasure also has one of de Paravicini's machines.
Kavi Alexander, the engineer with Water Lily Acoustics, won his Grammy for engineering a Ry Cooder album with a de Paravicini recorder; it was a 1-inch, two-track Studer C37. Alexander used a Blumlein stereo pair of custom microphones, which was built with rectangular mike capsules by Mitab and with tube electronics by de Paravicini.
When I interviewed Tim de Paravicini, I was struck by his strong, original opinions about audio.
--------
In some circles you have the reputation of a hi-fi tweak.
I'm not. I'm too academic to get into that.
The hi-fi fraternity is bizarre, full of dangerous amateurs. I try to steer clear and do genuinely innovative work-something that's worthwhile.
What caused you to start modifying recorders?
I was dissatisfied with their performance. If "line out" doesn't sound like "line in," that's not good enough.
What's the main advantage of your 1-inch analog recorder over digital recorders?
The sound quality. My analog recorder has four times the sampling frequency! The bias frequency is 160 kHz. The magnetic particle flow past a playback head is equivalent to a 24-bit word, which is amazing resolution.
Analog recorders can sound wonderful, but DATs are so portable and convenient.
Oh, God, I hate DATs. Stopping and starting with those things is a pain in the ass. With an open-reel tape, you can pause it and go instantly; it's human; it's tactile. Whereas DATs stop, fit, fart, and think about what they're going to do-they're just not friendly. And unlike DAT tape, analog 1-inch tape archives beautifully. Some tapes made in 1955 are earthshakingly good. They still sound fresh. But talk to anybody with U-Matic tapes or DATs, and see how well they store. Run a DAT through a machine 20 times, and you're struggling.
If analog tape sounds so much better than digital, what improvements should be made in A/D, D/A converters?
First of all, the frequency response should extend from 3 Hz to 50 kHz, because we experience those frequency limits. We are able to detect audio up to 50 kHz. We don't hear it, but we experience it in other ways. I can give you tinnitus very quickly if I run an ultrasonic cleaner at 45 kHz. You are aware that it's on, and your ears ring when it's shut off. On the low end, we detect mechanical vibrations down to 3 Hz. When a marching band walks past you, you feel the drums in your stomach and bones. And that's all part of the sound.
Ten years ago in Stereophile, I said that digital was never going to work well in the chosen format. Digital should use a 400 kHz sampling rate and 24-bit words. Then it will satisfy the hearing mechanism and won't have a digital sound. Digital has a "sound" purely because it is based on lousy mathematics. The manufacturers pre-suppose too simplistic a view of our hearing mechanism.
But manufacturers don't want to change--it's the lowest-common-denominator syndrome. It's like 525-line television, which allows you only X amount of resolution. With digital, you've fixed your resolution parameters, where analog never had that problem.
I still do work on the vinyl record; it still can be advanced. The number of vinyl molecules passing the needle every second is equivalent to half a gigahertz. So there ain't a lot wrong with it, fundamentally.
You can carry on improving it without losing compatibility. It's like good old 35-mm films-you carry on improving films, but there's nothing to stop you from shoving them through the same old projectors! I've been pioneering work on a CD player that runs at 88k, but it only works with CDs that were cut at 88k.
When storage density increases enough, we won't have the excuse for using only 44.1k.
Right. The manufacturers should have said, "Let's go gung ho and create a real system that works right." A 12-inch LaserDisc would have given you an hour's worth of music to the highest standard. Manufacturers try to pretend that what's good enough for Joe Doe at $5 is the state of the art.
Getting back to your recorder mods, what transport modifications do you make?
Whatever makes the system more stable. Actually, wow and flutter is more limited by the tape than the transport. Tape is a mechanically compliant item, and 1-inch tape has much more tape stability and strength than quarter-inch. Yes, the wider tape costs more, but it's a small part of the total project cost.
How did you first learn about tape-head design?
I stripped a lot of machines on the market, worked out my own mathematics, and read papers by Jay McKnight, from Ampex, and so on. I ask the tape-head manufacturers, like Saki, to build heads to my specific requirements. I emphasize bass performance and the fineness of the laminations.
Do you recommend using Dolby SR with your machines?
No. It's unnecessary, and it doesn't work well under dynamic piano conditions. It doesn't encode a control signal, so it can only approximate. The noise floor in my system is limited only by the microphone. You can't hear its tape hiss-just micro phone hiss.
You use vacuum tubes in many of your de signs. Some people have said that tubes have euphonic even-order harmonic distortion. Do you rely on this tube nonlinearity to achieve the sound of your mods, or do you always run the tubes in their linear region?
I do not rely on tube nonlinearity. I don't want a sound in my machines. What comes out must sound the same as what went in.
The "warmth" in a lot of tube electronics is due to their dismal top end, the bad transformers they use, and the loading down of their high-impedance outputs.
Because of the output transformer and the feedback used, many tube circuits have a partial bass instability that gives a bloated bass. Any warmth in the tube sound is a defect, but listeners don't want to know that.
I don't have to use tubes in my designs; I only do it for marketing reasons. I've got an exact equivalent in solid state. I can make either type do the same job, and I have no preference. People can't pick which is which. And electrons have no memory of where they've been! The end result is what counts.
Most transistor-circuit architecture was different from tube-circuit architecture, and that's what people were hearing, more than the device itself. The main advantage of tubes is that an average tube has more gain than an average transistor. Second, tubes don't have the enormous storage times of transistors, so they are very fast. Tubes go to 100 MHz without trying.
Moving on to microphones, your mikes use rectangular diaphragms, tube electronics, and huge transformers. Why?
A circular diaphragm has one dominant resonant mode. But a rectangular diaphragm does not have the same resonant mode in both axes, so it tends to have a flatter response. Also, a rectangular diaphragm has less off-axis coloration in the horizontal plane than does a circular diaphragm of the same area.
My mikes are transformer-coupled, triode designs. The electronics have a frequency response of 5 Hz to 35 kHz (-1 dB).
I use transformers in my microphones because they can do the job better than anything else. There's no advantage in transformerless circuits because a lot o them can't drive long lines. As long as I know that the electronics of my micro phone go from 3 Hz to 100 kHz at the end of 1,000 meters of cable, I'm all right.
Some transformerless mikes have pathetic headroom. Disgusting. We're besot ted with this phantom-power philosophy Most of the mikes draw only 1 damn milliamp at 48 volts, max. That's 48 mW o energy; it doesn't give you a lot of head room. I want a mike that can shove 3 volts +12 dBm, down a line, 20 to 20k, boom! Why so much voltage?
Suppose you take capacitor mike that produces 10-mV out put with 74-dB SPL input. At 144 dB SPL the mike will put out over 3 volts.
You've said that we experience sound down to 3 Hz, and that reproduction down to this frequency is essential. Do studio con soles go down that far?
No. The average console has all these cumulatively rubbish electronics in it. If you cascade 10 amplifiers, each with a response down 1 dB from 20 Hz to 20 kHz, you end up with a cumulative 10 dB down from 20 Hz to 20 kHz. So you must minimize all degradation. Since I use a lot of transformers in my stuff, each transformer must be very wideband.
Unfortunately, the average manufacturer looks at only one piece of equipment, in isolation. They quote a tape machine as having a response of 50 Hz to 15 kHz, ±1 dB, and say that's fine. Yes, in isolation. But not as a cumulative system.
Tony Faulkner uses a mixing console of mine, full of tubes and transformers, but it's vastly flatter than most of the mixing consoles on the market.
What's your overall design philosophy?
Audio devices should not have a sound of their own; they should be virtually a black box. The results prove themselves in recordings using my products. They do the job.
Whatever the device is, I look at it and say, can that device be logically improved? Forget about cost. Companies like Neumann charge a lot of money, but I say, could they make that product a little better and charge a little more for it? Try to make things better, whether it's outrageous or not. Somebody will want it and will pay for it.
Any last words?
I try to provoke people. I'm sick and tired of the me-too factor, the lemming factor. Just because everybody else wants to jump off a cliff doesn't mean I have to.
Many audio companies tend to rest on their laurels and don't bother to take the next step forward. They should leap ahead instead of staying on the back burner. They have the potential to be stunning.
Discography
The two recordings below, among the finest-sounding available, were made with de Paravicini modified tape machines and tube microphones.
Ry Cooder/Nishwa Bhatt: A Meeting by the Ricer. Water Lily Acoustics WLA-CS-29-CD. Kavi Alexander, producer. (Available from May Audio Marketing, P.O. Box 1048 , Champlain, N Y. 12919).
This beautiful blend of Western and Eastern acoustic music was Stereophile's Recording of the Month, and was reviewed in April 1993. In my opinion, the highs are especially sweet and gentle. The sound just flows effortlessly into your ears. The only hiss you hear, which is very slight, is that of the microphones.
Ustad Ali Akbar Khan: Indian Architexture. Water Lily Acoustics WLA-ES-20. Kavi Alexander and Sallie Reynolds, co-producers. (Available as a double LP from Finch & Marsh, 2457 Cascade Trail, Cool, Cal. 35614; 916/885-2279).
The following four recordings, again with excellent sonics, were mastered by de Paravicini and John Denz at The Exchange on a custom-built, all-tube cutting system designed by de Paravicini.
Ravel: Daphnis et Chloe, Boston Symphony Orchestra, Charles Munch. Chesky RC15.
The Power of the Orchestra/Mussorgsky, Royal Philharmonic Orchestra, Rent Leibowitz. Chesky RC30.
Clark Terry: Portraits. Chesky JR-2.
Ana Cararn: Rio After Dark. Chesky JR-28.
(adapted from Audio magazine, Jan. 1995)
= = = =