They are the workhorses of the film sound world, enduring endless days toiling in often hostile environments, arriving on the set before shooting begins and staying after it ends, there for every take and retake. They are the production sound mixers: location recordists charged with the responsibility of capturing crisp dialog and evocative ambiences; the pure performances. And though their work is often overshadowed by the more glamorous contributions of the post community, production sound mixers are the first critical link in the film sound chain. At best, their work is utterly transparent, not calling attention to itself. At worst, well, if you have to struggle and strain to understand what the characters in a film are saying, or noisy backgrounds overwhelm the dialog, or the finished film is overloaded with badly done ADR, chances are that the production sound mixer had a rough outing. And it probably wasn't his or her fault.
To gain some insights into different aspects of the world of production sound, Mix spoke with three top professionals in the field: Jeff Wexler, Mark Ulano and Glenn Berkovitz. All three are members of the Cinema Audio Society.
You might say that Jeff Wexler was born into film. He's the son of Oscar-winning cinematographer (and occasional director) Haskell Wexler, and he notes, “I absorbed a tremendous amount of knowledge from Haskell and from just being around the set all of my life, riding around on the dollies at the age of 2 and things like that. My first day on the set, when I actually had to do a job, felt very familiar; I felt like I'd been there all of my life, which I actually had been.”
Not wanting to compete with or draw comparisons to his famous father, he gravitated toward production sound, breaking in on “low-budget, non-union Roger Corman movies, where a lot of people got their start.” Now entering his fourth decade in the field, Wexler has amassed an impressive list of credits, including such films as Bound for Glory, 9 to 5, The Natural, Ghost, War of the Roses, Get Shorty, Independence Day (for which he was nominated for an Oscar), Jerry Maguire, Fight Club, Almost Famous, 61*, Rat Race and Vanilla Sky. When we spoke in early July, he had wrapped up work on Martin Brest's new film, Gigli, starring Ben Affleck and Jennifer Lopez, and had done some re-shoots (“they call them ‘scene enhancements’ now,” he says with a chuckle) for the Owen Wilson/Eddie Murphy film I Spy. Along the way, Wexler's been able to work on a number of films and commercials with his father; that's quite an accomplished team.
“For the most part, all directors proclaim that they're ‘really into sound; it's really important,’” Wexler says. “Then you begin to discover what they really mean by that, and you find out that the fundamental reality is that the entire soundtrack can be done later. And this is one of the things that separates the production sound department from almost all of the other departments on a movie set the day of shooting. Even if the sound team completely screws up, or the director completely screws up, you still don't lose a day of shooting. You can loop things, use ADR; in fact, they refer to ADR as ‘re-shooting the sound.’ Everyone knows this, and some directors are quite willing to say, ‘I don't care, get me a guide track, the sun's going down. We need to get it on film; I'll worry about the sound part of it later.’ That's the extreme position, although it's also sometimes the realistic position if you are working with a lot of mechanical effects or wind machines, where the sound we can record on the day could never be in the movie.
“Almost all directors say that they don't want to loop anything, and they say it for all of the right creative reasons, which is the importance of the performance on the day and that sort of thing. They all know it does not hurt the movie in terms of the budget or time to do most of the sound later. It's in every budget. The best situation is where you have an experienced and realistic director, a talented and experienced first assistant director, a location manager who understands that you cannot make a scene work if you're trying to shoot a three-walled set constructed under the freeway and it's supposed to be Iris' bedroom; it just isn't going to work.
“I've been extremely lucky to work with people who do have an allegiance to the production soundtrack because of all the things it can bring to a movie. Not just the fact that you don't have to replace the dialog. But the fact that the actual recording itself lends something to the movie; it tells you something about the characters. I've always said that I feel like I've done a really good job if I can sit in dailies with my eyes closed and listen to a scene — the raw recordings before anyone else has gotten to them — and you can tell something about the characters, how they're feeling, where they are in the room, what emotions they're having.
“Obviously, that doesn't happen all the time. A lot of times, I'm just glad I can hear something, because maybe we were fighting the wind machine, or the actor insisted on playing with his props during his whole speech and there are noises, or you have the actors who sound just fine when you're talking to them as real people and as soon as they start acting, you can't understand a word they're saying. Those situations aren't nice, but they happen all the time. I've had directors come to me and say, ‘Could you understand what she was saying? Is there some sort of microphone we can use to make that more understandable?’ I say, ‘Take off your headphones, forget about the microphones. Go over and listen to this person speak. Tell me if you can understand anything she's saying.’”
Wexler does a considerable amount of preparation before his first day on the set, talking to the director and feeling out his or her needs with regards to the production sound, getting involved in location scouting whenever possible and making notes on the shooting script: “I do an extensive breakdown on the script as soon as I get it and highlight areas where I think there might be some difficulty and areas that need some attention before shooting; whether it's building something somewhere, or changing something physically in a location. I might say, ‘Can we find a place that maybe looks the same but is a little quieter than the place you've chosen?’ Because if we're doing a quiet, emotional scene, you don't want to have to go to the actors and say, ‘You gotta speak louder.' Or, ‘We've got to get this quickly because at one o'clock, they fire up this factory next door.’”
Where technology is concerned, Wexler has always been a part of the vanguard in production sound, experimenting with new gear before it was widely used, but noting that “the changes that have come in our field, say, in the last five years, are all equipment-based technological changes that haven't had much impact on the actual work that we do. Recording sound is still not a big mystery; there's little magic involved. It still boils down to, if things sound good and you record them properly, they will sound good, and all the technology in the world will not change that.”
That said, he was among the first to use DAT on the set, and he was an early convert to the increasingly ubiquitous Deva nonlinear digital system. “When I first started using DAT,” he recalls, “I was not about to make any director or producer be the guinea pig for a new format that everyone believed would never work. So I ran my Nagra, which I completely entrusted all of my work to, and ran both machines for the first three pictures, and then I went to running two DAT machines [on War of the Roses in 1989]. At that time, I was a co-owner of the post-production facility where all of the sound transfers were being made, so I didn't have to let anybody know what we were doing. On the set, people would came over, look at the sound cart and see the Nagra and figure everything was fine.
“Then, when I started using the Deva, there was initially quite a lot of resistance from people in post-production and from various others. Of course, ultimately the promise of nonlinear work is that we'll be using the same sound files; there won't be questions about the quality of transferring the material because it's digital data, not audio.”
Though the Deva has 4-track capability, Wexler notes, “I almost never use all four tracks. That's my own personal choice. I'm still an old mono kind of guy; I'd rather give them one good track that makes sense to everybody. It requires no apology or explanation: ‘Oh, the good sound is on three or four.’ Or, ‘When you remix it, it'll sound great!’ There are other sound mixers who love using multitrack, and they also shower a scene with multiple techniques. In other words, they're not content to just boom the scene with one microphone; they'll also put wireless on the actors, they'll put plant mics. And whenever you get involved in multiple techniques, it would be suicide to commit all those to one track; you need to split those things off. You need to make them discrete so that somebody can reconstruct whatever it was you were thinking of on the day later.”
Wexler will use the other tracks for additional material that might be useful to the editors later, such as a sync ambience track or live effects that are essential parts of the scene, like a television playing in the room or a telephone ringing. But generally he likes to get as much as he can in mono off the boom while using wireless mics on the other tracks. Wexler especially trusts his boom operator: He's worked with Don Coufal for 25 years. “He really understands what the camera sees,” Wexler says, “and he understands what he needs to do to make dialog work. And that's really the most important thing that we do: Make the dialog work for the picture.”
Though the same fundamentals of recording location sound apply to both feature films and television, the pace of work on series television is faster and the budgets are smaller.
“Anybody who works TV will tell you it's definitely a very long day,” comments veteran TV production sound mixer Glenn Berkovitz. “The exception is to have something under 12 hours per day.” Berkovitz, who did production sound last year for the popular series Crossing Jordan, and whose other credits include work on such series as Freakylinks, Resurrection Blvd., Clueless, Chicago Hope, Lois and Clark and Dream On, as well as various TV movies and feature films including Zero Effect, Freejack, Young Guns 2 and Robocop 2, explains that “usually there will be a template for a show where you'll shoot eight days per one-hour episode. A lot of times, it will be five days studio and three days on location.”
In episodic television, the chain of command is usually different from the way it is in feature films because directors are generally hired guns for each episode, and the overall look, feel and sound of the show have been established over previous episodes, sometimes in the pilot. “Quite often, the individual episode's director is gone after shooting,” Berkovitz says, “and the machine of the show operates without a director once it gets into post-production. Editors cut the basic picture of the show per the style they've refined. Then the post-production department takes over: If the sound supervisor, who usually works for the sound house that is contracted to the show, thinks a scene is dicey from a sound point of view, or there are scenes that they think they can make crisper, they will spot those lines for ADR.
“It varies from series to series how much ADR is usually done, which usually happens about a week-and-a-half after you wrap an episode. They get an ADR list together and then they walk in the actors. The guest players usually don't mind, but the featured actors who are working every day usually aren't anxious to go over to the looping stage after work or during lunch. On Crossing Jordan, they started out looping a lot, but by the end of the season, the actors were less eager to accommodate; they would loop for technical fixes, not for performance. In features, they will sometimes loop a line for performance, but in TV, if you don't get the performance on camera, you usually don't get the opportunity to massage it in post-production.”
Berkovitz, who got his start in the business in live sound and lighting for rock 'n' roll, was a studio engineer before he moved into audio post. He transitioned into production sound full time in the mid-'80s, and has also worked on a pair of reality TV series: MTV's Real World (Los Angeles) and Making the Band, which traced the creation of the popular boy-band O-Town. This is grueling work in a different way: Generally, each of the principals in a reality series will have a two-person crew assigned to him or her — camera and sound — who unobtrusively follow the “actors” around on a 10-hour shift, stopping when the actor stops, moving when the actor moves. If there are four principals in a room together, there might also be four crews out of sight, with each sound person responsible for different coverage of the moment.
“Usually, you stick a wireless mic on each person, actually, they often put it on themselves; it becomes second nature. They wake up in the morning and put on a mic. And then you run a boom, as well,” Berkovitz says. “When I worked on Real World, it was occasionally difficult when you'd have several crews in a room together, but it's easier now that they've got some very good frequency-agile audio gear. Each actor gets their own frequency, and each sound person will have three or four receivers on their little walk-around sound rigs, and you dial it in. You say, ‘I've got actor B,’ and you touch B on your receiver and suddenly you've got their audio. It really worked well [on Making the Band], and that's now become the norm.
“I understand that all of Survivor is on booms, so the sound folks must get in as close as possible, gauge what their camera and potentially what any other camera, is seeing, focus on the action, but also know the camera might leave the action for something else. You've got to stay with the meat of the dialog, and then when the camera moves, get a coherent sentence before the audio leaves — so it can be edited with decent continuity — and then follow the camera. It's fun, but it's also hard. You have to think, and you have to be physical.”
For his regular production sound work, Berkovitz's gear of choice includes the Deva, with a Nagra ¼-inch as a backup, an 8-input Cooper mixer, Sanken CS-3 short shotgun mics and pair of Schoeps mics he uses for stereo: “I do as much stereo work as I can. Sometimes, it's just splitting channels for foreground and background, but if there's a street scene and it kicks into something that's not dialog — I don't really run stereo dialog — where I can get away with putting a stereo pair up there, I will.”
Of the Deva he says, “On the three series we've done with Deva, that choice was always driven by post, and I think it's an excellent way to go. The machine itself could be further refined: It's clunky. Physically, the box sits on the cart oddly and is somewhat fragile in its connectors. The ergonomics are not great, but it records very, very well and very reliably. They're constantly updating their software, which is nice. The post-production guys like it because they can load the sound files of our day's work and they don't have to do it in real time, and they can go back and check something immediately without having to shuttle through a tape. That's why we used it on Crossing Jordan.”
For a more specific look at how production sound technology is used in the field, we turned to Mark Ulano, a 20-plus-year veteran who won an Oscar for his work on Titanic, and whose lengthy résumé includes such notable recent features as Austin Powers, Jackie Brown, The Patriot, Stuart Little, Spy Kids, The Majestic and the recently released surf movie Blue Crush.
From the set of the next Quentin Tarantino film, Kill Bill, currently being shot in China, Ulano described some of the obstacles he faced during the making of Blue Crush, much of which was filmed at various beaches on the Hawaiian island of Oahu. He also described how recent technological strides are making his job as a production sound mixer easier.
“Blue Crush, which is being produced by Brian Grazer [A Beautiful Mind] for Imagine Films, had multiple creative challenges for the production sound crew,” he says. “Helping me solve these was my friend, master boom operator Tom Hartig, and a wonderful sound man from Oahu, John Reynolds, as our second boom/utility sound technician. We were faced with recording a great deal of dialog out in the surf, at the break. Most of our principal female actors rarely wore more than very skimpy bikinis, so that precluded much in the way of choices for body miking. Our primary approach for the surf work was via wireless boom on a combination of floating platforms, ranging from boats, jet ski and, most often, an inner tube with a suspension seat. John Reynolds developed this last technique during his time on the Baywatch Hawaii series. We also used a lot of the waterproof pouch/Aquapac technology we developed for use on Titanic. Some of Aquapac's newer pouches now allow for generic cable exits, making for more flexibility in our choices of wireless configuration. The boomer's rig mainly consisted of roll-top, waterproof neck pouches we got at REI that contained the microphone transmitter and a ComTek receiver.
“Our primary wireless,” he continues, “was either the new Zaxcom digital unit or the new Audio Limited Envoy. Both performed superbly under great duress, and I believe this was the first time a production model Zaxcom had been used for the whole show. Our primary boom mic was the magical Sanken CS-3, Rev.2. When we weren't actually in the surf and instead shooting on the beach, these Sankens gave us extraordinary control over what background noise was or wasn't in the pattern. This didn't always save us from discontinuity because of changing wave patterns in the background, but we were able to capture a much higher percentage than would be traditionally possible. We also captured a variety of stereo ambiences for different weather and wave conditions, which were intended to help smooth out the hard changes.”
Like the other mixers interviewed for this piece, Ulano has been using the Deva hard disk system, in part, he says, because “it is the first practical production field recorder to address the realities of post-production's digital realm. We can deliver the work in the resident file format that most of post works with — the media is vastly superior as far as archival stability is concerned — and we can multitrack, if needed, with no hassle. It's 24-bit, plus it's the pioneer for a whole family of devices only now coming to market.”
What kind of feedback did Ulano get from post people about the Deva? “It has been mixed, but it's changing,” he says. “The main bottleneck has been at the Avid, as they have been very slow to acknowledge the need to read their own SD2 file format. This is finally improving, as the pressure to move away from real-time import flows in from the customer base. There is also a conservative element in the studio post-production supervisor community, and as often as not, the sound supervisor isn't hired on early enough to engage in the discussion [about what format to use], so the decision often is more about familiarity than specifics. Approach is another issue. Many wonderful and talented post sound supers are very different in their styles, some are surgical in what they bring to the mix, preferring the editing process to be where elemental choices are made, and others want to bring every resource into the mix and work subtractively. Both styles produce great results, but they are completely different processes.
“I like to supply as many options to the post sound crew as possible, but only if it is in context to the work and, hopefully, the style of the post suite. What's the point of struggling for sync ambient stereo pairs to mate with the primary dialog/effects tracks if the philosophy of the post crew is to disregard that kind of material? On the other hand, if all things are in alignment and I get the opportunity to connect with our post people in pre-production, we can go through the script together with a fine-tooth comb and talk about preferences and hopes and so on. I had this opportunity on The Majestic with Mark Mangini and Richard Anderson [of Weddington Productions], and it made all the difference. I knew that they really wanted large stereo ambiences of those big crowd scenes with hundreds of extras in vast rooms. It wasn't a library item to them.”
Ulano is also a great believer in the new Zaxcom digital wireless system, which he views as “a great breakthrough. Zaxcom has eliminated the need for companding by using a purely digital system, thus affording the full dynamic range possible with digital technology. It's quite startling the first time you use it for a scene with wide dynamic swings. It's tiny, it's sturdy, it sounds marvelous and it has many innovative features. But it's also still early in its production life and, being software-based, new capabilities for it are frequently coming down the pike, in the Zaxcom tradition of intimate end-user inclusion for developing new stuff. There is a latency issue if you mix the Zaxcom with analog wireless equipment, but most of this can be controlled by mixing or going with a digital mixer, such as Zaxcom's Cameo, allowing for matching the delay channel-by-channel.
“While the ability to use wireless boom has been around for a long time,” Ulano adds, “it had been traditionally used as an occasional solution to a challenging shot, such as a 360-degree camera move. I've been treating the whole subject very differently since my move to Audio Limited UHF-diversity wireless technology. This was around the time I did Austin Powers and Titanic in 1996. I broke with tradition and committed what was considered professional heresy: using a wireless boom all the time. With the equipment performance levels that had become available, it became the only way to fly for me, and it was instrumental for our work on Titanic. Of course, this required working out a lot of issues, and it ultimately resulted in double-transmission from the boom operators, in the form of the primary transmission off the boom, and the PL/IFB secondary transmission to the private sound channel, which we run all day long; we use Lectrosonics for this. It also required a dual monitoring transmission: one public channel of program and one private one for our team and all of its communication needs. I use a pair of Comtek BST-50b's for this task. I wrote an article about this a few years back and was amazed by how many people contacted me to discuss my experiences. Now, I see the method appearing more and more as mixers become more trusting of their wireless hardware. It truly frees my crew from a host of unnecessary cabling issues and lets them focus on finessing their boom work and responding instantly to sudden changes on the set.
“Our biggest problem with the whole issue of wireless is the lack of frequency allocation. There is literally none for motion pictures or television. We are considered secondary users, and there's no movement toward improvement. As all the DTV channels are coming online, we are just leaves in the wind; just at the moment we most need to expand our frequency usage. I don't think this will be solved until a group of major players in Hollywood make that phone call pushing the political system for allocation. This melodrama continues to unfold. Stay tuned.”
Blair Jackson is the senior editor of Mix.