Interlude: Retracing My Steps
One of the questions I’m occasionally asked when I lecture is, “How do you feel about CDs versus LPs?” I can always tell the preference of the person who asks, because people who prefer compact discs never think twice about the matter; the question is asked only by people with a nostalgic love of LPs.
They’re always disappointed when I reply, “LP is definitely the best format—if you like pop, clicks, hisses and skips.”
When the CD arrived in the early 1980s, I was immediately taken by it, and realized that it would inevitably take over the marketplace for recorded music. The sound quality was obviously superior, the dynamic range was significantly extended and, yes, the absence of surface noise was a major advantage. Unlike LPs, CDs did not deteriorate with repeated use, because there was no needle gouging into the surface, only a beam of light playing over it. They were far more easily transported or stored. And they were incredibly cheap to manufacture: By the 1990s a CD could be manufactured for between 40 and 45 cents—and 30 cents of that was for the jewel case it came in. Marketing aside, the rest of the retail cost was profit. (In 1978 the retail price of an LP averaged $7.38, so a CD could substantially undercut that price and still make a handsome profit.)
Most of all, though, to a producer the CD represented space, and space represented time. The LP’s maximum of 30 minutes per side had been a quantum leap over the five minutes per side of a 45, but the 74-80 minutes available on a CD meant that we were in a new world. Suddenly people could listen to almost any symphony without the need to flip a record, and entire Broadway shows fit on a single disc without the crushing need to trim or even eliminate songs. In terms of what it could do, and how much it could hold, the CD was truly revolutionary.
In the first wave of stereo in the late 1950s and early 1960s, new music quickly migrated from mono to stereo, but the real boom was in rerecording old music, be it classical warhorses, Broadway favorites or pop hits. The same was true with compact discs, but with a major difference: To remake a mono record in stereo required a new recording, with all the effort and expense that entailed, but to rerelease an LP on compact disc required nothing but transferring it onto a CD, which any competent engineer could do in a matter of hours. The chance to make new money from old recordings was irresistible, and such transfers boomed.
To those who understood recording technology, though, an LP simply transferred to a CD was the equivalent of a mono recording played on a stereo: It played as well as ever, but no better—it failed to take any real advantage of the new medium’s technological advances. Recordings produced for LP used only a fraction of the CD’s capabilities, and it wasn’t long before listeners started to realize that the sound quality of new music on CDs was far better than that of older music rereleased on CD.
The answer to this was a process known as remastering, in which someone would go back to the original tape tracks produced in sessions from as far back as the 1940s and remix them (sometimes re-editing them as well) on modern equipment to produce new masters which would take full advantage of the capabilities of the CD. Soon a new generation of CD rereleases began appearing, featuring sound quality that was markedly superior to both the LPs and previous CD versions; they also often included “bonus tracks,” usually consisting of previously unreleased material from the original recording sessions. The packaging and promotion of these new rereleases typically trumpeted them as having been “Remastered for CD.”
Some of them even said, “Remastered for CD by Thomas Z. Shepard.”
People who don’t know the record business think of a producer as sitting in a control room, running a recording session. That’s a big part of the job, but the editing and mixing is at least as important. It’s a double-barreled job, and no producer worth his salt would hire somebody to run the session for him, or hand over the editing and mixing to anyone except a very trusted assistant.
Remastering a recording, therefore, amounts to redoing half of the original producer’s job, with the possibility of producing a very different recording. There’s nothing wrong with this, of course, if the remastering is being done by the original producer. In that case the producer is simply being given a newer, more powerful set of tools to revisit his original work.
But if the person doing the remastering is not the original producer, it’s quite possible that the intentions of the people involved in the original recording (including not only the producer, but also the songwriters, the author of the book and the director and producers of the show itself) will be drastically altered—and then released claiming to be a faithful restoration of the original album
(An extreme example: Someone remastering my original-cast recording of Song and Dance would have access to all the sessions tapes, including the instrumental half of the show which I eventually decided to omit; he or she might choose to edit, mix and release all that material, radically changing the recording as I left it.)
On July 5, 1986, I published an op-ed piece in Billboard which the editors titled “Remastering for CD: The Missing Ingredient in the Mix.” I wrote that very old recordings could only benefit from being made accessible to new listeners through new technology, and that very recent recordings were typically in the hands of the same people who had made them. Between these extremes, however, I saw a dangerous situation evolving:
“I refer to the remastering, perhaps the re-editing and most certainly the remixing of recordings made over the last 25 years by artists, producers or engineers who might be very much alive and available, but are not consulted for reasons of company politics or expedience.
“This issue is not theoretical; it is real and can be both personally painful and esthetically destructive.”
I called on the record companies to follow a simple rule: “If the original producer (or engineer or recording artist) is willing and available to assist in the intelligent re-creation of the artist’s original intent, then why not use him? If they’re not available, perhaps the CD transfer should be made from the original dual-track mix. I would prefer extra hiss to a mangled mix.”
It was ironic, therefore, that I spent a great deal of time in the 1990s remastering recordings, mine and other producers’, which had been released between the 1940s and the early 1980s. It was, if nothing else, a chance for me to practice what I’d preached.
The actual experience of remastering a cast album is very similar to the post-session experience of making a new one. Once again I found myself sitting in a tiny editing cubicle with an engineer, listening to session tapes and asking myself the same questions I always asked: “What am I hearing in this take, and how might it fit into the best possible recording of this song, this scene, this show?”
From a practical point of view, the most important difference is that, with the cast album for a current show, the performers are available to come back and repair any omissions or any mistakes that slipped through the cracks on the session. In remastering, the tapes or acetate recordings are all you have, and everything you do in remastering is limited by what is and isn’t on those original tracks.
The most obvious advantage I had was that I was under less time pressure than the original producers had been. There’s no arbitrary deadline for a remastered recording, so what was a matter of a day or two for the original release was typically as long as a week for a remaster. This let me listen to all the sessions tapes unhurriedly, and sometimes I came across very good takes that had been rejected or overlooked in the original session, or that had been unusable due to the state of that era’s technology, but could be rescued with current resources.
Beyond the practical considerations, the biggest difference between mastering and remastering, at least for me, is philosophical. As I’d written in my Billboard commentary, I felt that it was essential to respect the intentions of the original producer. If I was remastering my own recordings, well and good—I felt entitled to rethink my own decisions if newer, more promising ideas occurred to me during the remastering.
If I wasn’t the original producer, however, it would never occur to me to let my own aesthetic be my only guide. A good example is my 1998 remastering of Goddard’s original-cast recording of A Little Night Music. I had made my own recording of the London cast of this show only two years later, and it reflected my own very strong sense of how it should be recorded. I made no attempt, however, to bring that sense to bear in remastering Goddard’s recording; to the extent that I made any changes, it was because they seemed to bring the recording closer to what he was trying to do. A couple of time, I must admit, I overruled the original producer, which I regret today.
(For example, in the original-cast album of Gypsy, there was a minor trumpet glitch in the overture—a single trumpeter hit and immediately corrected a wrong note in the ending of the “You’ll Never Get Away from Me” section. Maybe Goddard didn’t notice it in his rush, or perhaps he didn’t have time to fix it. I felt sure that it would be consistent with his interpretation to fix it, so I did!)
There is, in short, a huge gulf between “Produced by Thomas Z. Shepard” and “Remastered for CD by Thomas Z. Shepard.” I’ve made a case for a Broadway producer’s work as an art form in its own right; I’d never make that case for a remasterer’s work, and I hope that no recording of mine ever ends up in the hands of a remasterer who thinks of himself as an artist in his own right.
I was doing my remastering for Sony Records, and it was my good fortune that the catalogs of both Columbia and RCA had wound up in the hands of Sony. This accounted for the vast majority of the recordings I’d made in my career to that time, and meant that I had a reasonable chance of doing the remastering for those recordings. Many of them were released on CD (nearly all of my Broadway output has been) without being remastered, but I ultimately wound up remastering 14 of my own original-cast and studio albums: Show Boat, Lady in the Dark, Annie Get Your Gun, Bajour, George M!, Dear World, 1776, Dames at Sea, Two by Two, Company, No, No, Nanette, Raisin, Sweeney Todd and Merrily We Roll Along.
The ones that I didn’t get to remaster myself have generally been in the hands of competent people. None has been ruined, and many of them have been done quite well. Nonetheless, there are always details that I’d have done differently, and I regret not having gotten the chance to do so.
An extreme example of this comes from my classical side, a 1967 recording of Debussy’s La Mer, with Pierre Boulez leading the New Philharmonia Orchestra. In those days the recording tape had a more restricted dynamic range. If something was too loud when you recorded it, believe it or not, the signal could bleed over into the next revolution of the tape, the way a drop of spilled water can bleed through three or four thicknesses of a roll of paper towels. Technicians in later eras weren’t usually aware of this, but we were conscious of it at the time, and compensated for it when necessary.
The very last sound in La Mer is a single loud, staccato eighth-note chord, played by the strings and timpani. The effect is that of one loud blow on the drum (you don’t really distinguish the sound of the strings, as such). The dynamic marking is fff in the strings and ff in the timpani; this means fortissimo, i.e. “very loud,” but we knew that, if we recorded it at its performance level, it would saturate the tape. So instead we recorded it more softly and, when I got to the end of the final mix for the two-track master, I brought the volume up to make it as loud as Debussy intended. Well, whoever remastered that recording didn’t realize that the soft chord on the master was compensatory—soft only because we needed to avoid that saturation. If he had consulted Debussy’s score, he’d have known what to do, but apparently he didn’t. All he knew was that, on the edited, unmixed dual-track tapes, the piece ended with a soft chord. So on the LP recording it’s a tremendous thwack, but on the remastered CD it’s little more than a tap. It annoys me, and I’m sure it would annoy Debussy.
Another one I didn’t get to remaster was Harold Sings Arlen (with Friend). Barbra Streisand had convinced Columbia Records to put her in charge of all her old masters. When she oversaw a remastered Harold Sings Arlen, it had been remixed, too aggressively for my taste. What I particularly regretted was that that on one track Peter Matz had wanted to add some additional percussion, so we did that as an overdub. In the remastering, though, they went back to the original three- or four-track tapes, and either they didn’t have that overdub or they didn’t realize what it was, so it’s not on the remaster. It’s a detail, but it bothers me. There are certain things you take the trouble to do because you know the recording in every detail, and it’s hard for somebody coming in fresh to understand them.
Sometimes the problem is not with the person doing the remastering, but the person doing the new liner notes. There’s a 1969 recording of Bach’s The Art of Fugue that I made with pianist Charles Rosen, and in one section the score calls for piano four hands, which for Bach would have meant using two pianists. Instead we had Charles do an overdub, playing a duet with himself. Well, as I’ve mentioned previously, one of the ways I entertain myself is by creating anagrams, and in this case I doodled on my session notes, “Second piano, Olsen Archers,” which is an anagram of “Charles Rosen.” I didn’t intend for this to go anywhere, and it didn’t … at the time. But, decades later, whoever did the remastering went back to the original session sheets, saw my note about Olsen Archers and, lo and behold, when the CD came out, the jacket said, “Second piano, Olsen Archers.” Charles had a fit; personally, I thought it was rather funny!
(I myself put one of my anagrams into print during the remastering process: On Goddard’s 1960 reunion recording of On the Town, a studio album featuring nearly all of the original cast with composer Leonard Bernstein conducting, there’s a song called “The Real Coney Island” in which you hear a carnival barker giving his spiel. The barker was played by Lenny himself, just for fun—he wasn’t credited and didn’t want to be. But when I remastered it, Lenny was no longer around and I thought we should acknowledge his performance, even obliquely. So I credited the barker’s spiel to Randel Striboneen, an anagram of “Leonard Bernstein.”)
Sometimes the changes you make in remastering are tiny ones, like not putting in that extra percussion or finally getting the busy signal in the opening number of Company to be on the right pitch, but sometimes they’re more substantial. When I remastered 1776 in 1992, I completely redid the editing and mixing for “Molasses to Rum”; even at the time I knew the original wasn’t the best I could do, but I’d simply run out of time—the masters had to be shipped to the factory. The remastered CD absolutely expresses my original intentions for that number; the LP does not.
The biggest thing you can do to make a remastered version more appealing than the original version, and thus to boost its sales, is to restore material cut from the original release, usually as bonus tracks. These numbers were almost always cut because there simply wasn’t room for everything on a 44-minute LP—but on CD you had another half hour or so to play with, so there was always room.
In remastering No, No, Nanette, I was delighted to restore “Peach on the Beach” and “Only a Moment Ago”; I’d never wanted to cut them in the first place, so it was great to be able to put them back. In Raisin “A Whole Lotta Sunlight” was a kind of musical scene for Virginia Capers, with singing and dialogue going back and forth. I’d cut out the dialogue and left only the song on the original-cast album, but I rethought that decision and, on the remastered version, kept the scene the way the creative team had intended it to be.
As I discussed earlier, I hadn’t edited and mixed Annie Get Your Gun (1963), because of a dispute over the credits with Irving Townsend, who wanted to be listed as co-producer. I was assigned the remaster, though, which meant that, three decades later, it was my turn to slog through the phrase-by-phrase syncing of Doris Day’s freely sung vocals with Franz Allers’ orchestra. That was really an ordeal, and at the end of it I put my co-producer’s credit back onto the recording. Townsend had died by then, so he wouldn’t mind, and I felt that I’d earned it.
The Sony catalog included far more than shows than the ones I’d produced, and I ended up remastering some of the true Broadway masterpieces: Goddard’s 1960 studio version of On the Town, with most of the original cast; Finian’s Rainbow; Bells Are Ringing; The Most Happy Fella; Brigadoon; Gypsy; Fiddler on the Roof; Mame; A Little Night Music and Annie. A particularly gratifying assignment was to remaster Goddard’s 1951 recording of Porgy and Bess, which had made such an immeasurable difference to my own life and work.
On every assignment I did my best to stick to the philosophy I’d outlined in my Billboard piece. Most of the original producers of these shows were no longer with us, but I tried my hardest to respect their work and ensure its ongoing legacy on CD. Occasionally I slipped, but most of the time I think I succeeded.
One of those slips was on Gypsy. When I started going through the tapes, I discovered an alternate take of “All I Need Is the Girl,” sung by Tulsa (Paul Wallace) to Louise (Sandra Church) as he tells her about the nightclub act he’s choreographed for himself and a female partner. On Goddard’s original album, Tulsa sings one full refrain of the song, then dances it for Louise as he talks her through it: “I start easy. Now I’m more debonair.” Near the end of the dance, as he says, over the music, “And now the tempo changes, and all the lights come up, and I build for the finale,” he continues to talk, but the orchestra swells so loud that we can make out his excited voice, but not what he’s saying, and the song ends with a big orchestral finish. It’s a very effective ending.
On the alternate take that I discovered, however, Louise impulsively joins in the dance, and you can hear her laughing with delight as Tulsa exclaims, “Louise, that’s it! Take my hand! Again! Again!” That laughter wasn’t on Goddard’s recording, and Tulsa’s final lines were all but drowned out by the orchestra. I thought her delighted laughter and the fact that she’s dancing with him were important for Louise’s character, because it’s the first time she takes pleasure in performing. It helps establish the roots of her later success as Gypsy Rose Lee. If I had been producing the original-cast album, that’s the take I would have chosen, so I chose that take for the CD version over the one that Goddard had originally chosen, and I mixed it with the orchestra still big, but considerably softer, so that the laughter is heard and the lines are clearly audible. And now, if you listen to that song online or on the remastered CD, that’s the take you hear.
Now, this was early in my years of remastering. If you asked me to remaster that recording today, I wouldn’t make the same choice. I knew that Goddard very rarely incorporated dialogue or sound effects into his recordings, and the existence of this one take with the laughter on it meant that Goddard had considered the same idea I did, but had rejected it. My choice, while valid, wasn’t what Goddard wanted. Today I would leave his original alone; I would include the track with Louise’s laughter, but strictly as a bonus track. The original would still be there, uncompromised.
But that’s hindsight speaking. At that time Sony had no particular philosophy about remastering. They said, “If you’re producing the remix, you do what you think is right, period.” So that’s what I did.
It’s not always possible to understand the original producers’ thinking. When I got to work on remastering the original-cast album for Fiddler on the Roof, which had been produced for RCA by George Marek and Andy Wiswell, I found two substantial pieces of music that had been recorded on the session but excluded from the album. One was “The Rumor,” and the other was the famous “Bottle Dance” or “Wedding Dance.” What I couldn’t figure out was why they were left off the album. It was not a question of time—the recording’s not that long, and they could have fitted them in. Somehow or other, somebody decided not to put them on the record. I never did work that out, but I put them both on the record as bonus tracks.
There were no missing songs to be restored to Finian’s Rainbow, but it was probably the most challenging remaster that I had to deal with, because it was the oldest. The original-cast album had been recorded in 1947, before the LP, before stereo and before recording tape, so producer Mitchell Ayres was limited to (by modern standards) very primitive technology … and it showed. One of the show’s big songs is “The Begat,” a rousing gospel-style number. The performance on the original release is fine—but, on listening to the various takes, I discovered that one of the other takes was substantially better. When I got to the end of that take, though, I realized why he hadn’t used it: On almost the last note of the song, there’s a really obvious mistake, which ruins an otherwise ideal take.
And in those days it was ruined, because with the acetate-disk technology of that day, it wasn’t possible to edit a take. You could only choose which take to use, so Ayres chose the best take that didn’t have glaring mistakes. For me, in 2000, that was no longer the case. Having converted the session acetates to digital tape, it was the work of a few minutes to pick up the clean ending from one of the other takes, splice it onto the really wonderful take and, presto, ensure that the remastered version would have a distinctly better performance of that one song than the original-cast album had had.
I don’t want to overstate the importance of the remastering process. I didn’t bring to it everything I bring to recording an album from scratch, and I didn’t want to. Even if I’d had a recording of all the dialogue from Annie, rendered by the original cast, and a massive library of sound effects, I wouldn’t have wanted to transform the 1977 original-cast album into a Tom Shepard-style radio play. I have too much respect for the original album, produced by Larry Morton and Charles Strouse. They made a brilliant recording, and it wasn’t my place to question the decisions they made. If the people behind Annie had wanted the original-cast recording to be made by me, in my style, they’d have brought it to RCA.
At most, the remastering producer is like an art restorer or a film preservationist, bringing his own knowledge and, whenever possible, the benefits of modern technology to a classic work and, through careful attention to detail, helping the original artist’s vision to live again. The worst thing a remastering producer can do is to confuse himself with the original producer, and I hope I’ve avoided that.
In remastering my own work, I felt free to be bolder—essentially, the idea of my remaster of 1776 was to do what I’d have done in 1969, if I’d had unlimited time and 1990s editing technology to do it with. In remastering the work of earlier producers, I felt that I was serving them, perhaps in the same way that, in my own recordings, I’m serving the creators of the show itself. The payoff, for me, came in large part in the opportunity to really dig into these classic recordings of legendary shows, to understand more than I ever had before the work of producers like Ayres, Marek, Morton, Wiswell and, above all, Goddard; and to more fully appreciate the work of the songwriters who had created those brilliant scores.
It was an interesting era, and one that, for a variety of reasons, is all but ended now. Nearly all of the pre-1980s musicals that were worth the effort had been remastered by the end of the 1990s, and today very few pre-1980s producers are alive to work on remasters, even if they were being made. The wave of remastering was driven by the change from LPs to CDs, and the most recent technological change—from CDs to streaming—has not produced any similar wave. What remastering is done today is almost always concerned only with updating sonics, and involves few if any creative decisions.
My time supervising remasters was very much a sideshow to my career as a producer, but it was an invaluable experience that, I’d like to think, made me better at what I do and improved the subsequent recordings I’ve made.
