Settling The Score
It’s easy to get caught up in the moment of any performance.
9 February 2013
Your version of what took place will always be framed by the interests and motivations you brought to the experience. For me, in the case of this production of West Side Story, I was drawn to the ability of the orchestra to stay in sync with the projection and spent much of the first half of the concert watching Ben conduct with the aid of the streamers. The difference between the space and tone of the MSO within Hamer Hall, by comparison to the original Hollywood sound stage was for me great enough to create a disconnect between score and image; making it harder to suspend disbelief. Following the interval, this barrier melted away and for much of the time I forgot the orchestra was even present.
Throughout the program it never occurred to me that I was also hearing something special through the house PA. I wondered, in the lead up, if we would be hearing a remastered surround mix of the film but aside from that I barely gave it a second thought. My experiences in the modern digital studio, allowed me to make assumptions, incorrect as they turned out, about how easy it must be to reassemble a version of the soundtrack minus the original score. Of course, you just dig out one of the saved copies of all of the different stems, that you’ve safely archived and hey presto. “If only!” I can hear them plead. I suspect it would take an entire issue of the magazine to do justice to the work of those involved but I’m very grateful to them for sharing at least some of their secrets.
EXCAVATION
When legendary sound mixer Murray Spivack convinced director Robert Wise to employ the new 6-track sound for the 1961 release of West Side Story, he played a major role in establishing a new audio standard for the 70mm format. Unlike our current surround formats, 6-track was a multi-channel stereo array with a single mono surround channel. The five channels across the front of the stage were assigned Left, Left-Centre, Centre, Right-Centre, Right.
While Spivack’s work was rewarded with an Oscar for best achievement in sound, it has not been heard by audiences for more than 30 years. A 20th anniversary remaster in 1981, created a new ‘6-track’ mix, using only a 4-track copy as its source (Left, Centre, Right, Surround). When MGM, the current copyright holder of the United Artists picture, began considering a 50th anniversary re-release of the film on Blu-ray the soundtrack was again in the picture. After listening to several reels of an original 70mm mag striped print held in the Academy Archive, Chris Lane, the Acquisition Manager for MGM Technical Services, was convinced that the superiority of the stereo separation for the orchestra, dialogue and effects warranted the restoration of an original 6-track mix. Only one problem… they couldn’t find one.
After an exhaustive 12 month search, through the MGM records, interviews with Spivak, and some old fashioned digging through the shelves, they had unearthed more than 100 individual reels for Chace Audio by Deluxe to appraise and digitise. Chace itself had been involved in post-production servicing for West Side Story for over 20 years so its own records also proved critical. As Chace Audio’s Bob Heiber told AT, “For the restoration we had an incomplete set of 14 reels from the 6-track English master from 1961 (mistakenly labelled, February 21, 1978); we used portions from the 6-track ‘Minus vocals M&E’ (Music & Effects), which comprised a complete set of 21 reels, as well as the most original 4-track LCRS composite mix (21 reels). We later found a second set of 6-track mags, original to the 1960s, to make up the complete set. All told about 42 reels of mag were digitised in the 6-track or 4-track format.”
RESTORATION
Bob Heiber: “Our mag transfers were done using the OMA-SE Chace from Sondor (an archival grade master recorder with a modular headstack system to facilitate the reproduction of a full range of magnetic and optical formats) into Pro Tools at 24-bit/96k using the custom Plangent Clarity Electronics to capture the bias information necessary for the Clarity Non-uniform sampling algorithm to do its magic.”
Take a breath… and a bow! The acetate cellulose mags were suffering from Vinegar Syndrome which causes them to shrink and the edges to curl. Through a collaboration between Chace and Sondor the OMA-SE was modified, including a custom headstack, to minimise the affect of these physical deformations and limit the symptomatic wow and flutter that usually results.
Working with researchers at Cambridge University, Plangent Processes developed their Clarity Audio Restoration. This research revealed that much of the distortion typically attributed to tape machine electronics, is perhaps more accurately identified as the result of the mechanical instability that endures almost unavoidably at the heart of the machines’ designs. Wow, flutter, scrape flutter, bass cancellation, muddiness, grainy midrange sidetones, interstitial haze and transient blurring all contribute to a reproduced signal differing significantly from the original source. Clarity recovers signals in the ultrasonic region that can be found on tape (bias tones… etc) using a proprietary Unix-based engine, running at 768k with 32-bit double precision processing. These tones are then tracked and ascribed the properties of a moving clock, similar to a variable sample rate. By stabilising these signals to a constant rate and processing all of the audio in the same way, Clarity essentially retimes the recording and in doing so recreates perfectly pitched audio (depending on the original performances of course!) with greater perceived depth and fidelity.
After transfers were completed, the Audio Cube AC-5 was used to correct all of the more typical restoration issues like hiss, pops, dropouts, edit bumps and distortion. Mastering was done in the THX-certified Rick Chace Theatre by Chris Reynolds (Chace’s Mixer and Technical Operations manager) with Lane from MGM. A ‘five across the front’ array was set up under the screen, using Dynaudio BM15A active monitors, to ensure that all channels could be isolated and scrutinised. While not used within the orchestral project a 7.1 surround mix was also made by James Young using DTS Neural Technology in ProTools.
DISSOCIATION
It’s one thing to have a beautifully restored film, but another thing entirely to turn it into a backing track for some of the great orchestras of the world to accompany. At this point in the process Chace handed off the broadcast waves of their six-track remasters to Audionamix. Head of Production, Rick Silva, was happy to spend some time explaining how Audionamix’s ADX reverse engineering technology all but erased one of the most famous orchestral scores ever recorded; just don’t expect him to give away any trade secrets.
Rick Silva: “ADX started as a research lab based in Paris, working on technology that could separate both musical instruments and melodic voices from typical mono or stereo music recordings for respatialisation within new multi-channel mixes. After our expansion to Los Angeles in 2010, ADX’s Music Dissociation service emerged following repeated requests from major studios to remove only the music track from a final composite mix, leaving the dialogue and effects tracks completely intact. This complex problem required the development of sophisticated algorithms by our Paris-based R&D department. It also demanded the skills of an experienced production team with a deep understanding of traditional audio engineering principles. Put rather simply, we gather specific sonic information about what we call the ‘known source’ (music that is to be removed from the mix) and our proprietary algorithms ‘learn’ it, so it can then effectively be removed from the mix.
“For West Side Story, we were given a stereo mix to use as the master for ADX’s Music Dissociation. The learning sources provided were a six-channel M&E with no vocals, the restored six-track mix with various degrees of lead vocal bleed in the individual channels, and a Left/Right from the six-channel M&E. Our instructions for this project were pretty straightforward — remove the orchestral score without removing or damaging any of the dialogue, effects, or melodic vocals that accompanied the score. When sources share similar tonal qualities in addition to overlapping harmonic and melodic content, the separation process becomes extremely difficult. The most challenging part of West Side Story was separating multiple vocal lines from the orchestral score that were performed in unison. To overcome this challenge, ADX provided at least two levels of separation. In areas where there was too much similarity in tonal and harmonic content to provide pristine separation, we supplied additional alternate versions that contained a small amount of the orchestral score content. This ensured that the vocal content sounded full and undamaged when played back with the live orchestra.”
RECONSTRUCTION
Once the orchestra had been dissociated from the master mix, the team at Chace Audio began working with the production team at the Hollywood Bowl to produce a final Vocal & Effects track for the premiere performances. As Bob Heiber recalled, “We did a gazillion edits to comp the best vocal stem from the supplied extractions, often opting for bleed over lack of intelligibility due to digital artefacts.” Through each performance, tweaks were made and the input of the house engineers was also incorporated. By the time the work premiered at Avery Fisher Hall in New York it was decided that more significant reworks were still required. The much-massaged ProTools project now found itself in Sync Sound’s Digital Cinema, New York’s largest mix stage, where engineer Ken Hahn was joined by Eleonor Sandresky (Associate Producer) and Garth Sunderland (Senior Music Editor) from The Leonard Bernstein Office. Both had been deeply involved in the ‘film with live orchestra’ project from its inception.
Ken Hahn: “The track needed to be rebalanced — smoothed out. This was accomplished by good old fashioned manual gain manipulation; accomplished with a combination of automated processing in ProTools; using Waves SSL G-series EQ and limiter, for high and low pass filtering, equalisation and comp/limiting. I used a Waves W-43, for general purpose hiss and noise reduction, McDSP notch filters for hum and extraneous tones and various Izotope plug-ins for selective note and noise elimination. Since the track we were creating had to be mixed live with an orchestra in a concert hall, I felt that the original film mix was too dynamic, and would require the live mixer to ride the track more than what would be expected. Some parts were too low, some were hot, some too dull, some too bright. The combination of EQ/filter and automated dynamic control, supplemented by lots of manual gain riding, produced the results we were looking for.”
In addition to remixing, the presentation of the material also changed, with the amplification of the dialogue and effects separated from any in-house reinforcement of the orchestra. Mike Runice was also added to the touring production to ensure that there was always an engineer/technician present with an intimate knowledge of the program.
RESPONSES