Doug Sax: From Direct-Cut to Compact Disc

As reported by Michael Fremer on, legendary mastering engineer and co-founder of Sheffield Lab, Doug Sax, passed away on April 2. Doug had been suffering from cancer and would have been 79 on April 26.

Coincidentally, we had just posted J. Gordon Holt's October 1982 review of the Sheffield Track Record, which Doug had cut direct-to-disc. This reminded me that Robert Harley had interviewed Doug in the October 1989 issue of Stereophile; rereading that interview reminded me that in September 1984, I had published an interview with Doug in the magazine Hi-Fi News, which I edited at that time.

So, in tribute to Doug, here is my 1984 interview, reprinted with the kind permission of Hi-Fi News editor Paul Miller.—John Atkinson

Record-cutting engineer Doug Sax is one of the founders of the Sheffield Lab record company and responsible for some superbly natural sounding direct-cut discs, as being eloquently opposed to digital recordings in general, and Compact Disc in particular. To quote Doug from our November 1983 issue: "...if the CD master was made from a good analog tape, then the LP can blow the CD out of the water" and "...I am not going to waste time reciting the litany of high fidelity rules that are being broken by the Compact Disc." Now Doug is a man with an impressive track record regarding sound quality and his disc-cutting facility, The Mastering Lab, is in much demand from artists and producers wanting the best sound quality from LP. Some have said that this involvement represents a vested interest which explains his lack of support for digital, but I have always felt that, as a man involved 24 hours a day in the assessment and production of high quality sound, his views have been worth taking seriously.

It came as some surprise, therefore, while visiting the June 1984 CES in Chicago to discover that Sheffield Lab were introducing a range of Compact Discs, made by JVC from the safety master tapes recorded during the direct-cut sessions for a number of their titles. I met with Doug Sax at the CES, therefore, and before talking about CD, I asked him when he had first had doubts about the digital encoding of music?

Doug Sax: It was at one of the first digital recordings ever made. Soundstream wanted to bring a machine down on one of Lincoln's sessions [Doug's Sheffield partner, the pianist Lincoln Mayorga]. It was being done with a single tube stereo mic over maybe eight strings, two oboes and everything; I had nothing to do with it except that they came into my facility to hear the digital tape against the analog tape that was run simultaneously. We put the thing up and I heard some things that I liked, but after about 10 minutes, I heard things that I had never heard before—the hall ambience, the lack of extended top, the lack of being able to listen into the instruments.

I didn't think anything else of it at the time, I didn't know then that this was the coming thing, that the world was going this way. I was just critical of what I heard. In a way it was a similar experience to when I heard my first solid-state amplifier back in 1958. Those amplifiers are now legendary for how bad they were, but at the time this man in a hi-fi store was thrilled with them and asked me if I had heard anything like that before. I said "Honestly no, not live, not recorded, not in my imagination! It's terrible." He said "Oh, you've been used to the softening effects of tubes and transformers." "I play in an orchestra," I said, "I hear violins warming up every day and that, my friend, is not a violin."

Obviously digital has come on since then, that early Soundstream was not a full-range machine, but in every case that I have listened to I am aware of things that I feel are wrong. Naturally there's things wrong in analog too, but the things wrong with digital are more of a musical nature; they get in the way of me reaching out towards the artist, towards the instruments. In preparation for CD, on product which I thought would have a broad appeal, I have run a digital tape simultaneously with the analog recorder and the direct disc for maybe three and a half, four years. I have been able to hear these machines under conditions where I knew what was going into them, and my misgivings about their sonic superiority were realised. And the better the signal I could produce, the more complicated it got, the more I felt these machines suffered.

Even on the CDs that I'm transferring from my analog tapes, to my ear I'm getting better overall musical results than I am on the digital. Now that may go against logic, because surely a direct digital transfer is going to show up its virtues better. But I know what my ears here, and one thing that I have learned is never to argue with them. I don't take the logic function that says "Well, this can't be happening therefore my ears must be wrong." because the ear is always right!

Atkinson: If the listener is honest to himself, that is...

Sax: Right. I think that as a general rule, people are so much in awe of any new technology. And the truth of the matter is that new technology works flawlessly to a fault except when it comes to audio. For instance, if you were going to go out and buy a home computer, and a knowledgeable man said "There's one coming out next week that has four times the memory storage, it will handle three floppy discs rather than one, and it's "100 times cheaper," you would get it and you would get exactly what you expected. In all areas except one, it always works. Radios, toasters, microwave ovens, you are told that this is the "new improved model," and it turns out to be exactly that, the "new improved model."

So with all that positive reinforcement, when it comes to audio the man is told that this is the new model, and it eliminates this and this and this, and it's better, he says "Right." But it's not so.'

Atkinson: The thing about digital is that the new technology is assumed to work perfectly: the math is right, therefore the machines must work without a flaw. And if you say "Hold on, I know it should work, but I'm glad when it is turned off," you're assumed to questioning Nyquist and Sampling Theory, not the anonymous engineer in the depths of Philips or whatever, who put together the practical realisation of the maths...

Sax: This device is not supposed to be something that feels good when you turn it off, it's supposed to be something that feels wonderful when you turn it on...Time is the absolute essence of evaluation, not what you hear on an A/B test. You don't listen to music on an A/B switch. If something's good, I get an emotional involvement. If I find that I can't get an emotional involvement, I want to know why? How come I can't get an emotional involvement in this? It sounds OK.

And when someone comes out and says "I have nothing to base this on, no scientific tests, but why am I not getting the thrill from this new technology that I routinely got from my old technology?", the bottom line is "Did or did it not add something meaningful to your life?" If it didn't add to your life, then you don't have to explain it. If I've been out with a beautiful girl and I introduce her to you and you go out with her and it just doesn't happen, I don't say "What do you mean it wasn't happening? The girl's bright, educated, she's her own woman—you tell me you like bright, educated women, what do you mean it didn't happen." You don't have to explain it, you just don't buy it.

The CD is only the film. All my efforts as an engineer have been in the camera and the lighting and hopefully where I put the actors. The CD's a film, I can argue, and you can test to show that this film is not as good as another format, that this new 16mm is not as good as the old 35mm. That's not the bottom line; the bottom line is that the world's great music, the world's great orchestras, and the world's great conductors, what, to someone who is into music, is the finest thing that has been achieved by man, is being stored in a manner that is unacceptable, or if acceptable, only marginally so, to a substantial amount of listeners.

Atkinson: So, how come Sheffield Lab are introducing CDs?

Sax: The greatest virtue of Compact Disc is that the user gets an X% level of performance all the time. 100% he may not get, but he doesn't need any operator knowledge—What turntable? What pickup? Is the overhang okay? The convenience aspect of CD is important—switch it on and it works. I'm not going to take sides in the sense that I'm going make my product available in both formats. If you go out and buy it, you'll know that each is the best I can do in that format and you make your choice.

I'm gonna make better-sounding CDs. I'm gonna run my lathe, I'm gonna run my analog tapes, and I'm gonna run a digital recorder, and I'm gonna get into that digital recorder because there is much going on there that I can fix from an analog standpoint that will make it better.

What I would like to see done, and I have a feeling that what you would like to see done, is when the companies that have the wherewithal to record the finest soundsources in the world—the Berlin Philharmonic, the London Symphony—put up a simple mike, position the damn thing, stick it on an analog tape recorder, and transfer right from that two-track to the disc. You tweeks out there think that's the way to go. At the same time, I'll go on with my mikes, I'll put it down on digital and make a Compact Disc that'll thrill the world.

I'm gonna get two markets for the price of one, that's why I'm looking at Compact Disc. The market wants Compact Disc and they want Sheffield product in that market, and they're getting it. But there's no way that I'm gonna take my Compact Disc in to any knowledgeable listener and convince him—I walk over into the high end at this show and I am half in fear of being assaulted. What bothers me with CD is that I can't get into the music. On the [James Newton Howard and friends] recording [CD-23], the pulse isn't right, the subtle movements aren't there on CD. And this group is the heart of Toto, they're great musicians, noted for feel. The direct disc I'm immediately in with the music. On one track "She," it's the piece de resistance of the album, the ending builds and the instruments layer, it expands. Now this is the high level where CD is supposed to be better, but it just packs up and gets uglier sounding. On the analog, it just blooms out and justifies musically what they did, the thing opens up and it just blossoms. The direct disc eats it up. And I would love there to be nothing wrong with CD. I wouldn't have to worry about LP cutting problems, no returns, no defective discs. I could make my mikes as good as I could. But it's not getting the job done. You're gonna have to explain to me why, if it's so perfect, the CD doesn't sound better than the LP?

Let me tell you something. When I was growing up, I knew by ear that London records were the King! Now they've sold off their analog machines. What I would say to them is don't sell them, bring them out of the back room, and record in parallel and get an extra market. If they feel it would be too small a line for them, I'll distribute it, I'll master it, they make the tape and I'll do the rest. And if there's no-one left at the record company who knows how to place a single stereo mike, then I'll come over and place the mike. I'll even provide the mike!

In other words, I'm gonna stay in business; they're gonna stay in business; CD may or may not go—I'm gonna make money on it while it's here—but why are these orchestras not being stored properly. That's the real crime.

A man will say that CD is ruining classical music. And I will say "What are you complaining about? You can't buy any kind of classical record these days that isn't made with digital. That's the problem." In essence, an LP record made from a digital tape is the worst of both worlds: if there is anything wrong with the digital process as it's being used, that low level of resolution's going on to tape, and in addition you have all the analog problems—warped discs, the ticks, the flutter, and increasing distortion with increasing level; the worst of both worlds. You'd just as well buy a CD.

I know that there are LPs cut from digital tapes that sound better than the CD, but I don't think it will be that way for ever because they are learning a lot more about transferring the information. In working with digital, I found a number of things to be true. Yes, you can't make an inaudible copy, and if you run a tape through a digital editor, the result is chop suey! Everyone uses editors so that they can make a digital fade and get out the little noise on the end; they're all concerned about not having the noises on their tapes, yet they never seem to be concerned about having music on them. When you go through a digital editor, even though the levels are fixed the information has to be rewritten and the loss is traumatic. You go in apples, you come out oranges.

You say that the advantage of digital is that master tapes can be stored without degradation. That's theory. Now I know of digital masters that have been used a lot which don't retrieve anymore, and we're not talking 30 years later, we're talking a year later. I'll tell you this much: if Decca would let me go in and grab some of their fine old analog masters, I'll guarantee that I can cut a record from it that'll make audiophiles get on their hands and knees and thank everybody that it exists. I get some old tapes in, done in the late '50s, and I'm not saying that they sound as good as they did when they were made, but some of them sound intimidating.

I mean intimidating...I've heard vocal reproduction, I've heard tapes where I'd be hard put to come up with anything today that sounds that good. I don't care if it's accurate, it's the way I like to hear it. You hearing more than you think. You're hearing analog; you're hearing all tubes, in the recorder, in the microphone; and you're hearing in the part of American tape, that's been stored much better than anything we make today, analog.

There's no question that there's one end of the chain which, if it isn't right, will mess up all your listening tests all the way back, and that's the power amp. I've gone in places where I couldn't tell any difference between things which I know should be dramatically different, and I can take the same tube devices under test into my studio with my amp—I've been with valves forever— and it's like night and day.

I would continually hear more resolution in the tube mikes than I could get out of the same capsule powered in any way that I know of solid-state. But I don't know if I would have that perception if my power amp were solid-state.

... on my own records I use my microphones— what's mine about them is the electronics...I'm fortunate to have a brother is arguably the finest tube designer in the world. The capsules are mostly the old AKG C12, though I do have some with Neumann and Schoeps capsules. What I have found since I've been working in sound—and all my efforts have been off the floor— is, forget about the storage medium, I've only got one lousy mic in here, a piece of wire and an amplifier. It should blow me away, and it doesn't; why doesn't it sound any better? Well, it must be the diaphragm, they all have a little peak around 12kHz, so I change that, but it still doesn't get right. What does get it right is when you get into the electronics, and I continually end up with these old valve microphones.

Atkinson: it's not just the mikes that are important, it's how you use them. Your Michael Newman guitar record [LAB 10]..."That's all tube; even my cutting amp is tube!"...I feel to be one of the most realistic recordings of a guitar in a room.

Sax: I have great respect for the way things sound live. I sat down in front of Michael for half an hour to hear exactly what the instrument sounded like. I was shocked. "Boy, it's not very loud! And it's so tiny!" 12 feet back and it's just this little sound source. It really sounds mono; you get a little back off the walls, but it's not even loud enough to excite the room very much. What am I gonna do? Am I gonna record it the way I know most people would prefer it, a little bigger, a little wider, or am I gonna present it the way it really is? I tend to go for the honesty, the resolution, and hope that some people recognise that.

What I hope to be doing soon is recording some major orchestras in analog, direct to disc, and with tubes. I just haven't had the money since the first two orchestral recordings. But with the Compact Discs, I'm gonna have the money...That's the bottom line: CD will generate enough income to enable me to do what I really want to do, which is to record Symphony orchestras in analog!

Allen Fant's picture

Excellent article- Thanks! for sharing.
PLease feel free to list your fave Sheffield Labs LP/CD (s).

hollowman's picture

JVC XRCD of Sheffield Labs material were not released in US, but they were in Asia (JVC Hong Kong or Korea ???).
I have some of those, including Sheffield Drum Record.
It's, in many ways, better than my vinyl copy of that album.
Whatever format, that recording is best-sounding drum acoustic I've heard (to date).

remlab's picture

Wow! Sax had some major cajones to stand up for what he heard with digital. Even when most of his income was derived from it. Excellent interview! Thanks, John for posting it.

John Atkinson's picture
"CD will generate enough income to enable me to do what I really want to do, which is to record Symphony orchestras in analog!"—Doug Sax

Doug, of course, did eventually record more symphony orchestras in analog. See

John Atkinson
Editor, Stereophile