A Tale of Two Studios
Preserving America’s musical heritage by transferring the RCA Living Stereo and Mercury Living Presence master-tapes to SACD by Richard S. Foster
AVAILABLE IN ISSUE 40

As an avowed fan of both the RCA Living Stereo and Mercury Living Presence records I’m fascinated by both their musical integrity and the passions they provoke amongst collectors and the owners of re-issue houses alike. Whole forests have perished in the discussions of relative qualities, while the computer time expended on internet debate must have materially contributed to the hole in the ozone layer. But with the arrival of the latest digital incarnations of these recordings, now transferred to SACD, the whole furore kicked off again. Would the silver disc finally banish the black? Would the bottom fall out of the market for mint condition, second-hand vinyl pressings? Would all those avaricious record dealers finally get the just deserts of their price-gouging past?

Of course, things are never that simple and what became abundantly clear at an early stage (and in a bizarre reversal of the reigning vinyl status quo) was that the RCA tapes had been significantly better served by the SACD re-issues than their Mercury equivalents. The longer I listened the more obvious the disparity became. The longer I listened the more frustrated I got. What, I wanted to know, accounted for the difference in quality between these discs? At this point I doubted it was down to the tapes, which left the transfer process itself or the pressing of the discs as the likely culprits.

In February 2005 I made arrangements to visit John Newton of Sound/Mirror in Boston. Sound/Mirror is mastering the RCA Living Stereo tapes for release on SACD, including the new down-converted CD layer. John (who is also the remastering supervisor) generously opened the doors to their studio and arranged for me to spend most of the day with himself and Mark Donahue, the mastering engineer responsible for the RCA Living Stereo SACD project. Dirk Sobotka, their DSD engineer was, unfortunately, not available. I would like to sincerely thank John and Mark for the generosity they displayed with their time and open access to their studio. My thanks also to Blanton Alspaugh, of Sound/ Mirror who single-handedly opened my ears to multi-channel surround sound with the Beethoven project he is currently working on. Blanton, I’m a believer. I hope to discuss this listening session at a later date. While attending the Munich High End show in May of this year, Kai Seemann of Speakers Corner Records arranged a visit for me to the Emil Berliner Haus in Hannover. Although most of this visit was spent with Willem Makkee and things analogue, Seemann was also able to arrange that I spend some time with Mr. Andrew Wedman, tonmeister (balance engineer) at Universal Music and the man responsible for the analogue to digital conversion on the Mercury SACD project. His time was extremely limited and I am truly appreciative of the kind attention he gave me while I was there. Thank you, Andrew. The time spent with Makkee as well as Seemann and my trip to Pallas will be saved for an upcoming article. Stay tuned, please.

The day I spent in a nondescript building near the corners of Green and Myrtle streets in Boston, was one of the most informative days I’ve ever experienced. The credentials of the people I was involved with at Sound/Mirror are impeccable: Newton was a staff technician for Dolby Labs and Vanguard Records before founding Sound/Mirror in 1972. Donahue is a graduate of the University of Massachusetts-Lowell, with a degree in Performance/Sound Recording Technology and has been with Sound/Mirror for well over a dozen years. He’s handled virtually every engineering seat within the organization. Alspaugh brings to Sound/Mirror many years of experience as a conductor in addition to his work as a producer, engineer and manager in commercial classical radio.Although unavailable this day, Sobotka received his Diplom-Tonmeister degree from the Hochschule fuer Musik, Detmold in 1996. Before joining Sound/Mirror he was a mastering engineer at Sound Byte Productions in New York City and some of his credits include working for labels such as Nonesuch Records, Teldec, EMI, Sony and BMG. In 2000 he was the recording and editing engineer of the first commercially released DVD-Audio containing a 192 kHz stereo layer.

In a nutshell, these guys know about cutting-edge digital sound and know where that edge is. But what’s equally impressive is their depth of knowledge regarding the history of the stereo recording process and technology, from early two-track tape releases in the 1950's through vinyl playback right up to where digital technology is now – and where it might be going. All this information was dropped on me before 11AM and believe me, I’m glad so much of it was recorded.

I also got to hear master-tapes that hadn’t been played for at least 45 years. How can I be so sure of that? The spools were secured with a particularly virulent type of splicing tape that hasn’t been used in 50 years – partly because of the foul deposits it leaves behind on removal. It was a salutary lesson as to the very real problems facing any serious remastering project on a purely practical level. It was there in Boston and it was there in Germany. In fact, it accounts for one of the most common issues with older tapes, along with shrinkage or stretching, poor splices and shedding. These tapes are deteriorating – in some cases at an alarming rate, quite literally fading or crumbling away – and the opportunities to preserve them are becoming increasingly limited. That’s the primary pitch for DSD, as it can be used comfortably as an archival medium once the analogue signals are converted to high-resolution digital and then stored on a hard drive. Sure, analogue copies would be lovely but dream on…

My main interest of course was the process itself; what happens from the time you place a master-tape on the machine through to its being replayed and converted to DSD. Except that I soon discovered that a huge and vital part of the process occurs before the tapes are even played. I hadn’t appreciated that just checking the master-tapes is such a major production. Simple winding and rewinding of the tape and checking for some of the problems we’ve just discussed is not an easy job. Then there is the manual alignment of the original tape with the heads of the tape playback machine being utilized. After that you must determine the proper equalization, etc. Once this is done, there is an initial listen to ‘see what you have’ and then there are decisions that need to be made if there are drop-outs, splicing issues or other unforeseen problems. It’s not a simple process. If you think this is just put the tape on the machine and push play... think again. What was common to both facilities is the amount of thought and problem solving effort required before they can even start. What you don’t want to do, in any way, shape, manner or form, is inadvertently cause any damage whatsoever or accidentally do something that cannot be reversed. This is NOT easy work. These were all things we discussed before physically going into the equipment room and actually performing the transfer tasks.

I asked the team at Sound/Mirror whether if at any time do they, or have they, listened to the original LPs. As far as Donahue is concerned, and I think he’s absolutely right, “the tapes are the Holy Grail. After all, the LPs were made from the tapes. A lot of the quality of an LP is euphonic from the distortions that are inherent in the playback medium, and I don’t see how you can garner more information from a medium that ultimately contains less information than the source material. When you listen to the tapes, it’s virtually impossible with a very, very wide dynamic range tape to get more information on the LP than exists on the tape. A lot of times, the cutting engineer was working very, very hard trying to get as much of what is on the tape onto the LP. There were decisions and compromises that were made at the time.”

Newton also makes some very good points. “Cutting lacquers was an art. There probably weren’t, at any one point in time, more than a dozen people who could cut a good classical lacquer – forgetting about the limitations of the machinery they were using. Understanding the music, understanding the dynamics of the music, and the limitations of the medium and making everything come together as good as possible.” In addition he adds, “More importantly, the producers and the artists of the time when all these decisions were made, were not thinking about the LP medium and its limitations. They were listening to the three-track output from the console and three-track playback from the tape. So, their original ideas are contained and determined that way and we wanted to hold those up as the main and most important value set, not what most people grew up with, the vinyl medium with its attendant issues.”
I learned that as regards the Living stereo tapes, with the exception of the 1954 sessions almost everything was recorded with three track machines. The exception was the recordings done in 1961 at Decca in London, which were strictly two-track sessions. Keep in mind that there were also mono and stereo teams working at the same time in those early sessions. So you could very easily have different mixes for the mono and the stereo releases. Approximately 1959 to 1960 is when this stopped and notes would then be added to the tape boxes, i.e., Do this for monaural and this for stereo, etc. Being a bit of an archaeologist also helps, believe me.

As an aside, we got into a brief discussion about cutting engineers and why there hasn’t been much written about who did the cutting at RCA. Many of us are aware of George Piros of Mercury stereo fame and John Johnson the cutting engineer of the Mercury monos. I learned that RCA was of course, a union shop and for many of these guys, it was just a job over “X” hours. In fact, you could have one person cutting the lacquers for an opera and when his shift was over, someone else would come in and finish the job. This could easily explain why some operas do not sound the same for all six or eight sides. As an interesting aside Newton talked a little more about how difficult it really was cutting quality lacquers because it was such a mechanical process with so many variables. A fair analogy could be the quality of an automobile manufactured on Wednesday versus one manufactured on Friday afternoon or Monday morning... Which would you want to own?

We went into the equipment room with the miles of cables, connections and a plethora of hard drives, analogue-to-digital converters (and vice versa) and computers, and I saw the 1980 Studer tape recorder, which has of course been highly modified and rebuilt. To this they added an Aria playback set. It’s a five-channel box: Two of the channels are optimized for the quarter inch stereo playback head and three of the channels are optimized for three channel half-inch playback head. There are two head blocks and of course these can be changed based upon what is being played. When they change the heads they change the channels that are in use because all the cable links and resonances are compensated for inside the box. The signal path comes out of the Aria and can be routed through an AME equaliser – which they had to make. They use this for AME tapes only; if the tape is not AME, then the signal goes straight from the Aria through the mastering chain into the A-D converter. Siltech cable is used for the entire signal path which is actually quite simple and straight forward. Everything is controlled remotely from the studios (I saw and heard three different studios, each room being a different size and each contained different speaker set-ups).

Mark was winding a tape and believe me, this is an art in itself. It’s done very slowly because of some of the problems we discussed earlier with the condition of the tapes. For example, because of the sticky tape issue if you do this too quickly, you lose information – which rears its head as a missing (or misplaced) chunk of oxide. The tape he was working on highlighted yet another problem; the calibration tones were missing.

The tones are typically at the head of the reel. This was done before the recording session had been started by the recording engineers. They calibrated the machines, for example for high and low-frequency preemphasis, bias and level. Basically what this meant was that at the beginning of each session they would tweak the particular machine to maximize its ability, establishing the replay parameters for the tape to be recorded. Without them it’s impossible to accurately replay the tape, regardless of its condition. Mark selected one he knew had the tones and said, “Listen to this.” First I hear a tone and then a voice comes on and says, “Disregard previous run. Use this run (then a tone) December 8, 1958 frequency Chicago, AME 1KC Stereo for level set.” The voice was that of Lewis Layton!

We then walked through the equipment room and into the mastering studio with its computers and mastering tools. Sitting on the desk was a switch box. Mark spent time adjusting levels as we were going to hear the material in three different formats: the original three-channel master tape, three-channel DSD and finally, two-channel DSD. All that would be necessary would be for me to push a button on the switch box to go from one format to the other. In front of the desk was a set-up consisting of three 1990's vintage B&W 801 series III speakers driven by a Threshold S500, the center-channel powered by a Classé amplifier. After Mark finished his set-up, he turned his chair over to me and said go for it! I sat at his desk and the tape began to roll. What immediately impressed me was the sheer brute power of the orchestra coming from all three speakers. A wall of sound! It really was as if I was sitting right in Orchestra Hall. You could hear everything. Certainly I’d never heard this much information from any vinyl playback source! It was truly fabulous. I hit a button and immediately I was listening not to the original tape, but hearing the tape through the DSD converters in three-channel sound. Again, phenomenally impressive and hitting the button back and forth to A-B the original tape directly or through the DSD converters proved a very interesting point: There wasn’t a whole lot of difference. Actually, let’s be more precise than that; the differences were so small as to be musically insignificant. Could I reliably identify the difference between the master and the DSD feed? No way. Okay, so there’s a switching box involved but even so, this is a pretty impressive performance – particularly for a digiphobe like me! I was, quite honestly, astonished. I really expected to hear a much wider shift between the two sources. Going to two-channel DSD was more obvious, but that’s got more to do with the replay chain and available headroom than the DSD format itself. This was still master-tape playback I was listening to and although it’s only two of the three channels available, it remained immensely impressive. This went on for almost twenty minutes and I can tell you I was quite literally shaken with what I heard. The Mahler 4th was wonderful – and not only the original tape – but listening through the DSD converters as well.

But most interestingly of all, the character of what I heard at Sound/Mirror with the RCA material, is exactly the same sonic character I get at home. Of course we’re talking different set-ups and all, but there is no manipulating of the signals as far as I can remember. They don’t fool with the goods. I remember what we listened to in Manchester last year while demonstrating good three channel SACD playback and we used the last movement of the Munch St. Saens’ Symphony No. 3 and once again, this is what I heard in Boston. The dominant sound of the original tapes is clear to hear, and what is on the tapes finds its way onto the Living Stereo SACDs – which might help explain why the three-channel replay of those discs is so darned impressive.

I had a great time with these exceptionally knowledgeable people and it’s reassuring to know just how capable the hands responsible for these transfers really are. I’m personally delighted they’ve been given the opportunity to do this work for Sony/BMG, breathing new life, digitally speaking, into not only the Living Stereo SACDs, but also the down-converted two-channel CD layer as well. For my money, it’s the best I’ve ever heard these performances sound from silver disc. Now I understand why.

It was some three months later, that driving through the outskirts of Hannover with Kai Seeman, we came upon the rather large complex that is Universal Music. After passing through the security gate we were expected at a small location neatly tucked in one corner of the grounds: Das Emile Berliner Haus! I knew the minute I met Andrew Wedman I’d like him. He’s Canadian. Joking aside, he’s a very nice man and totally devoted to what he’s doing at Universal. He was right in the middle of transferring The Music of Leroy Anderson, volume 2 (U.S. SR90043). While with Wedman, I asked many of the questions I had previously asked at Sound/Mirror and the answers were just as I expected. Probably the biggest problem they have is the condition of the tapes. Not only is there dramatic acetate tape shrinkage (some of the 35mm film tapes have shrunk in excess of two feet!) but there are problems with certain sections of material actually missing. For example, he mentioned that one of the future releases was going to be SR90054, the 1812 Overture. It took him forever to find the carillon part of the tape, and despite searching far and wide has yet to find the cannons portion of the recording. Clearly there have been gremlins in the archives, exacerbated by the convoluted history of Universal’s many acquisitions. With tapes having been moved from various parts of the world, shunted from pillar to post, things have clearly gotten more than a little mixed up. The audiophile view of master-tapes as the Holy Grail, cherished and preserved for posterity receives yet another serious dent, adding further to the need to archive this material now.

Because Andrew was literally in the middle of a transfer, I had neither the time nor the freedom to listen that I’d enjoyed at Sound/Mirror. But I was able to enjoy Anderson’s Typewriter piece. While Universal has all the toys and equipment as well as the bells and whistles to boot, I’m afraid that what I heard in the digital domain at Berliner Haus, is exactly what I hear at home from the Mercury SACDs. It’s missing some life and drama, transparency and dynamic range. I didn’t hear the master-tapes at this time, only the DSD feeds, but I heard enough of them earlier whilst I was with Willem Makkee (as well as knowing what he’s cut to vinyl for Speakers Corner) to know what’s out there as a source if you choose to use it. The sonic problems seem to lie in-house, because there’s no question that what I was hearing in the studio once again bears the same sonic thumbprint that characterized the first batch of Mercury SACD releases.

Discussing this with Andrew he showed me examples of the tapes he’s working from. Whereas Makkee is relying on 60’s vintage stereo mixdowns, the SACD issues are being drawn from the original three-channel masters. Now in theory that constitutes an advantage, but the chaos demonstrated by the mislaid elements of the 1812 combined with the actual physical condition of the tapes themselves more than negates the potential benefits. Andrew showed me one tape which had not only shrunk by more than three feet, but had succeeded in twisting back on itself through a full 720 degrees! The state of the oxide layer can only be imagined. I think he has been struggling manfully with the challenges presented by these tapes, which have demanded considerable work and even editing to make them useable. It’s a frustrating situation in which he finds himself, especially as the two-channel tapes are delivering demonstrably superior sound. What price the original master…

Incidentally, speaking of editing, the situation regarding DSD is far from straightforward and requires dedicated workstations. Sonoma and Sadie are the approved options, although widespread concerns regarding loss of signal quality (see the article regarding the Peter Gabriel SACDs in Issue 24) have meant that many people actually resort to editing in PCM before final conversion to DSD! However, there is now a dedicated 24bit/384kHz PCM based system called DXD (Digital eXteme Definition) which allows greater flexibility and capability without any loss of signal quality. It’s also interesting to note that both Berliner and Sound/Mirror had full suites of Meitner, Genelec and dCS converters, different units often being employed for different functions within the chain. But both studios independently professed a marked preference for the dCS stack when it came to classical work, so the digital hardware employed certainly doesn’t account for the differences heard. Whilst there are other equipment variables, such as the actual tape machines employed, one has to conclude that, after all it’s actually the tapes themselves that are the biggest variable.

However, what is clear from my experiences is that the DSD format is capable of exceptional performance as an archival medium. Its ability to accurately hold the information content of mastertape over time is far more stable than the analogue alternative. Indeed, it’s the failings of the latter that create most of the problems when it comes to the current preservation of this material. How successfully you can overcome the challenges presented by these crumbling artifacts in part defines just how good those archival transfers will sound. But the other part of the equation is how you carry out the transfer itself. DSD stores the information you feed it with self-effacing honesty and really can deliver on its promised performance. But like so much else in hi-fi it’s a case of garbage in, garbage out. Just rejoice in the fact that the tools are most definitely out there, if we choose to use them…

Top | Back to Last Page