Bernie Newnham
Alan Taylor has written some fascinating stuff recently, explaining how he got over problems with sound in different situations. We have his permission to wrap it up into a page on the Tech-ops website to save it as a little piece of valuable history.
Alan says that you can do all that stuff easily now, and it’s true, but that doesn’t detract from the abilities of innovative people to get a job done in the many years before “two clicks and it’s sorted”.
In the early days of non-linear editing we used to make multi camera recordings of sound checks, the biggest being at RAH on three Sony VX1000s. The offline editor we had was Eidos Optima, which only had one video track. How we would have loved to be able to do what you can do in the latest version of DaVinci Resolve in a couple of clicks.
Alan Taylor
My technique for dialogue replacement can now be done digitally in an instant, with syllable matching software making any delivery fit the original, or even make different languages look tolerably in sync.
I was the Sound Supervisor on “Life and Loves of a She Devil” and if I had to sum up working with Phillip Saville using just one word, it would be “challenging”, in a good way and also a not so good way.
When we shot “She Devil…”, the very first scene we recorded involved an absolute swine of a complex tracking shot. It started outside of a house, the camera tracked indoors and the actors moved from one room to another, delivering a lot of dialogue as they went. It was all done on a single shot, which presented all manner of challenges for the camera operator (Mike Winser), lighting (John King), myself and the boom ops and of course for the actors too. After we got it in the can, I mentioned to Philip that it was a hell of a difficult shot to attempt on the first afternoon of shooting. He smiled and said that he liked to see what the crew were capable of when put under pressure. We wrapped very early that day because the whole sequence was in the can and no additional shots were needed.
During the dubbing process, Philip made it clear that he despised the rather bureaucratic way the BBC operated when compared to commercial TV practices. He especially hated having to stop the dub at 21:00 and resume the next day if he felt we were making good progress and were on a roll. I would happily carry on a little longer, but we soon got hassled by engineers or security people and had to stop. There was one sequence where Philip was getting into a strop because we couldn’t source a suitable background sound effect of an American hotel lobby (for a sequence shot near Derby). Whatever we came up with wasn’t good enough or not American enough. When we packed up that night, I phoned an ex-Kendal Avenue engineer who I knew was then working in the BBC New York bureau and asked him if it was possible as a favour for a local recordist to make four or five atmosphere (“atmos”) recordings of different New York hotel lobbies and get them over to me ASAP. He readily agreed and arranged for them to be sent overnight via Concord ( it was a daily matter to send stuff that way, so it was just bundled in with what would have been sent anyway ). He put on some impressive stickers on the tape box saying “Urgent via Concord” and they were waiting for me in Television Centre (TVC) reception by the morning. Philip was impressed to say the least, and he absolutely loved the recordings. I pointed out to him that there are times when the size of the BBC and the fact that parts of the BBC operate in different time zones can also work to your advantage.
With regards to the shooting of “Life and Loves of a She Devil”, there were some interesting problems which we faced when shooting.
One of the principle locations was Mary Fisher’s house, which the book said was a high tower overlooking the sea. The location manager, Simon Moorhead, was unable to find a location which looked the part while still being suitable for shooting. The solution adopted was to use the Belle Toute tower near Beachy Head and to build a massive set around it to provide the other features and in particular the huge modern lounge with picture windows overlooking the sea. (See below for more trivia about Belle Toute).
When we started shooting in the living room set, there was a huge problem for sound because the set was effectively plywood cladding over a scaffolding frame and the exposed situation high on the cliffs meant that we were experiencing strong winds on a lovely sunny day. The wind was blowing through the scaffolding and sounding like the polar wind Fx (Effects) disk which was much used by grams ops. The sound wasn’t especially loud, but certainly audible. The issue with it was that we were indoors and there was nothing to explain that wind noise because everywhere looked calm. Even looking out to sea, the height of the cliffs meant that the visible parts of the sea were a couple of miles away and looked pretty smooth, even though it was horribly rough.
We tried all sorts of ways to discover exactly what was causing the whistling, but it would have wrecked the scenery to get inside it to investigate or treat it, so we were stuck with it. The solution I adopted was incredibly simple. My main concern was that the wind noise was unexplained, so I asked the art department if they could put some pot plants on the balcony so that we would see them blowing in the wind. Once some potted conifers were put there and being seen to be blown very hard by the wind or even blown over in shot, the fact that we could hear the wind was no longer a major problem.
Another challenge was that in the story, the She Devil character, Ruth (Julie T Wallace), undergoes extreme plastic surgery in America to look like her love rival, Mary Fisher (Patricia Hodge).
Obviously there is a process of transformation as Ruth starts to look more and more like Mary Fisher, but there comes a point where Patricia Hodge has to play the part of Ruth in her transformed state. The problem is that their voices couldn’t be more different and it was going to be essential to re-voice all of Patricia’s dialogue with Julie’s voice.
I worked on a children’s drama where a similar problem was faced because of prosthetic full face makeup making the actor’s voices unusable and I had devised a way of re-recording the dialogue on set after everybody else had gone home. On that children’s drama, the original dialogue was recorded on a timecode Nagra in the usual way (the snoop tape) and the director was invited to select no more than two or possibly three good takes for each shot. Those good takes were then copied to another timecoded Nagra and made into loops with an audible header edited in with three beeps at one second intervals, some loops included feed lines. The actor would then wear a wireless headset and listen to the looped dialogue while moving around the set as before, repeating their lines. I recorded the boom onto a second timecode Nagra which was also being fed with the timecode from the loop. The director was listening for a good performance and I was listening on split cans, with the original in one ear and the new dialogue in the other. If I was satisfied that it was in sync, I gave a thumbs up and if the director was happy with the performance, she gave a thumbs up, at which point we stopped the tape. I then used a red Sharpie pressed against the tape as I rewound it to the point where I heard the beeps at the start of that take and then wrote the scene and shot directly on the tape, which conveniently had a matt white backing. The VT editor subsequently reported that it was doddle to get those tapes to automatically sync to the original VT and then use the replacement dialogue throughout the edit. Our workflow on site was very efficient and in post-production it needed hardly any effort at all.
I mentioned to Philip Saville that this would be a neat way to solve our problem, but he was totally unconvinced and wanted to re-voice in a studio during the post production process, while I felt that it would be difficult to satisfactorily match the acoustics of the location sound in a studio. I sensed an opportunity when there was a scene at a funeral. Ruth had by that point completely transformed into the Mary Fisher lookalike, so Patricia Hodge was playing transformed Ruth sat in a Rolls Royce. She wound down the window and simply said to Bobbo (Denis Waterman), “Hi Bobbo, it’s Ruth … your wife”. We shot the scene with Patricia performing it and while they were setting up the next shot, I persuaded Julie to sit in the Roller and repeat the lines to an audio playback. We didn’t have any means of automatically syncing up 1/4″ tape machines on site, so resorted to trial and error, hitting the play button at the right moment and recording it to the spare VT track, or another VT – I can’t remember which. When we got it right, I asked Philip to look at the VT and he was very impressed to hear Julie speaking Patricia’s lines just half an hour after we shot them and sounding exactly right for being in that car. He immediately agreed that we should re-voice all of the Patricia / Ruth dialogue on location. After those scenes were shot, myself and a boom operator spent a couple of days revisiting those locations with Philip and Julie to re-record all that dialogue. It all dropped into sync so effortlessly at such an early stage of the edit and sounded so natural that Philip sometimes forgot that we had ever re-voiced them.
There are doubtless hundreds of Philip Saville and She Devil stories, but I offer an additional one about Philip and the voice overs at the end of each episode. This one is less technical.
Each episode ended with a highly dramatic voice over by Julie playing the She Devil, in which she relishes the evils which she had bestowed upon those who wronged her and then hinted at what she planned for next time. We recorded the voice overs in the sound studio adjacent to Sypher II and Philip had a clear vision of the immensely powerful delivery which he required. Julie was a wonderful and versatile actress, but I had to agree with Philip that were weren’t getting the performance which we knew she was capable of delivering. Eventually Philip declared that the problem was the studio fluorescent lights and what was needed was moody lighting with most of the studio in darkness. I found a couple of Anglepoise lamps and put a scrap of green gel over one lamp, illuminating her face from below. Julie could see her evil reflection in the studio window while she performed. We minimised the lighting level in the Sypher suite so that Julie was largely unaware of us and could see her reflection better. Philip approved the lighting and Julie ultimately delivered a performance which exceeded our expectations.
It gave me a golden opportunity to wind up John King by later pointing out that we even had to resolve lighting problems in the voice over booth, but the truth is probably that the improved delivery was nothing to do with the lighting, but more down to having a few moments respite for Julie to be fussed over and get into the role before continuing, however the exact reason didn’t matter because we got the result we needed. Personally I didn’t like the way that Philip insisted on the voice being quite so massively over compressed in order to make it super-loud whilst the closing music swelled behind it, but I had to admit that her sinister voice made the hairs on my neck tingle when it was transmitted. I later discovered that Philip was delighted to observe that the She Devil closing voice-over was so loud that presentation announcements after the programme were rather quiet by comparison.
======
Endnote:
Belle Toute: Some of you may recall that this tower was subsequently moved inland due to coastal erosion threatening it. A further interesting aside – the tower was in the process of being sold around the time we needed to shoot there and the vendor didn’t want to remain responsible for it, but the prospective owner couldn’t complete in time. The BBC ended up buying the tower and owning it for a short while until the original purchaser could complete. The lighthouse windows now seen on the top of the tower were not there previously, having been built for the shoot. The incoming owner asked for them to be left in place afterwards.
Pat Heigham
Totally fascinating, Alan, and a great example of ingenuity in solving a problem. As befits a BBC trained engineer.
I would say, however that the system of sync beeps or clicks was utilised on a movie I worked on: “Fiddler on the Roof” in 1971.
I was keen to utilise two track playback for the picture, having been brought up with two track playback at the BBC (Orchestra and Vocal separate, so the relative levels could be adjusted simply on the pots), avoiding the need for lots of tapes with different ratio balances, as had been the norm. I wasn’t about to make up loads of tapes with different ratios! But we needed a sync machine. Jim Willis was engaged in modifying an existing stereo Uher machine to run with sync-pulse lock, but abandoned that when news of Kudelski’s stereo Nagra was announced. Approaches to Switzerland only elicited the answer that it was then at least two years off and when it happened, it: “would use half-inch tape!” From the time schedule, this was a no-no, so Jim set about creating two half-track recorders from existing Nagra 4.2s. Special two-track heads were made by Branch & Appleby with a wider than normal guard band so that the existing Nagra Neopilot sync track could be accommodated. I helped him build the second channel record and play amps. They worked brilliantly, for Jim was a true engineer in the sense of being ‘ingenious’.
As Mirisch, the production company who made “West Side Story” was very wary of our intention to use ¼” tape, having done that previous musical off gramophone turntables driven with three phase linked to the camera drive. We considered 35mm magnetic film for the playback tracks, but the unwieldiness of spooling back ready for another take, and even if we went to 16mm, not a lot faster, they needed to be convinced that tape would work to the extent that a stage was booked at Pinewood, with camera crew and cast, for a day’s test. Only when they viewed rushes the next day did they acquiesce to our advice. Subsequently it was proved that spinning back the tape was quicker than the camera crew getting set up for another take.
We enjoyed a long prep time to get everything ready – ordering up all the kit required, making the playback tapes etc. and I made every single cable myself, XLR mic, PA speaker, and the 3-phase power cables for the cameras – yes, the camera dept. in those days, took the view that if Sound wanted it all to run in sync, then they should be responsible for the power supply!
Whether this was a retaliation for our department shutting the ‘noisy’ cameras into a booth, I don’t know, but it could have been!
“Fiddler on the Roof” – a major playback situation, made complicated by the need to record live dialogue in between the mimed verses.
I was intending to use line-source speakers from the dead side of mike, like we did in the BBC, with any leakage being covered by the laid in music tracks at the dub. However, the Music Editor explained that that wouldn’t work as the music tracks and the final dub would be stereo, and three-track at that.
So, live dialogue pick-up was achieved by utilising ‘silent’ playback with an induction loop run around the set, the actors wore induction receivers with moulded earpieces. The long hairstyles helped hide these! When we could use loudspeakers, I opted for a number of line source type spaced around so that on a long tracking shot, the actors came away from one and onto another, so they were not too far from a speaker at any time, and the sound level could be kept lowish.
They were Philips line-source speakers driven with 100v line and were pretty directional. The amps were Grampian (don’t laugh!). They could be powered from 240v AC or 12v battery. As the sound truck, fitted with a compressed-air mast carrying the aerial, also housed the shortwave radio comms to the production office back in the city, any time there was a need to talk to the office, the playback system had to be muted, as the radio signal induced into the amplifiers, and blasted out to the countryside!
The Music Editor, who had cut “West Side Story”, was supplied with a remote stop/start/dim control for the playback Nagra, as he stood close to the camera, intently watching the accuracy of the artistes’ lipsync and would cut the take if unhappy. Clicks were provided for the players to know when to start miming and close their mouths at the end. I made up the ¼” tapes from 3-track 35mm copies of the master music recordings – vocal/orchestra/clicks, plus a load of cassettes for the artistes to practice in their hotel rooms before the number was due to be shot – it could be many weeks between the original recording and appearing on set.
One afternoon, preparing for the processional lead-in to “Sunrise, Sunset”, Dick Carruth, the music editor arrived in the sound truck, which I had had customised fitted out, in a state of agitation.
“Norman (Director), wants 8 bars cut shorter for the procession! – it’ll take three days for me to get my UK assistant to make the edit, get a ¼” done and fly it out to us!”
Me: “When does Norman want to shoot?”
“This evening!”
I quietly took down a spare tape from the rack, asking Dick where the cut was required.
Razor blade and splicing tape in action – “Is that what you want?”
“Gee! I didn’t know you could do this on location!”
Brownie point to the Brits!
Alan Taylor
The system I devised for on-site dialogue replacement used influences from many areas and I was always open to adapting promising ideas from anywhere. I was very fortunate during my BBC career for having an unusually wide range of experience within the BBC, far beyond the formal training at the time. Although being recruited into the engineering side (TA course 30), I had actually wanted to be a sound recordist, but didn’t know the proper job title when as a schoolboy I wrote to the BBC saying I fancied a job as a sound engineer. The BBC latched on to the “engineer” word, and once I got to Wood Norton I realised it wasn’t the job I intended. I managed to end up at Kendal Avenue and in those days, you pretty well stayed where you were sent and would rise through the ranks wherever you were based. Moving sideways to Tech Ops was viewed as demotion by my line manager in the Engineering Department, so I slowed down my progression through the engineering ranks by seeking attachments to every other department that I could think of. I did camera operation, radio links (comms), video tape operation and even stage management, but crucially did several attachments to the OB sound dept. This was long before new recruits were sent on a tour of different disciplines.
After I eventually transferred to sound and rose through the ranks to become a Sound Supervisor, I unexpectedly found myself specialising in drama, but had a nagging feeling that there were better ways of approaching OB drama rather than being based solely in the truck. I eventually informally approached the Film Unit at Ealing and was warmly encouraged by Roger Turner and invited to see how they worked. He in turn was fascinated to find out more about how, in video drama, it was the responsibility of the Sound Supervisor to record the sound during shooting and also to operate the Sypher dubbing suite in post-production. I was enthusiastic about how being responsible for pre-production, location shooting and also post-production audio gave me unique advantages and it meant that my knowledge of the dubbing process could facilitate aspects of the location shoot, while being on location meant that the location recordings could be optimised for post production. Visiting film shoots demonstrated that there was a lot in common between how we worked on OBs and on film, but it also gave me a very early introduction to the Ursta Cart, which proved to be the key to escaping from the sound control room in the Scanner in order to work on the set. A timecode Nagra served as the snoop recorder while a battery powered 6 channel mixer, or even an SQN mixer was used, finally a multi flail allowed me to hook up via a multi-cable to the Scanner to send stereo audio both ways along with timecode. The need for communications via the Scanner became minimal as there was no need for sound talkback and otherwise, the main requirement was to run VT and to report when it was at speed. Working adjacent to the director meant that communication was face to face, rather than via the First Assistant via a walkie-talkie and I usually had visual contact with the boom operator, so verbal communication was either easier or superfluous. Once I got that system perfected, I rarely worked in the sound end of the LPU unless there was a particularly complicated sequence with multiple microphones. This way of working is quite commonplace now, but I started working this way from the mid 1980s onwards. I was surprised and delighted by the encouragement given to me by Roger Turner as he had just become a line manager in a rival department. At that time, most location drama work would either be shot on film via Ealing or via Kendal Avenue, so in effect he was assisting a competitor. He was also a strong advocate of M-S stereo recording and gave me custom built decoders and monitoring boxes so that I could experiment with M-S at Kendal Avenue.
I was especially pleased that my workflow for dialogue replacement proved to be very efficient, but also only used standard equipment which was already on site and in editing suites. I knew that it was normal practice in VT to use a lay-off 1/4″ tape recorder which was timecode synchronised with the VT machine, so it was obvious to me that providing replacement dialogue which appeared to be conceptually the same as a lay-off recording was going to make life easier all round. Writing take numbers directly on the tape and having a broad red line drawn through all the required takes meant that it was easy to find the takes. My first experiments revealed that life was impossible for the tape synchronising controller when the timecode had multiple instances of the same timecode, therefore by visibly marking the tape, it could be rapidly be set up at the playback head ready to run up to sync as soon as the VT has spooled to the next cue. It soon became obvious that it had to be spooled manually near to the appropriate point as each unwanted take also had the correct timecode so it couldn’t be spooled and parked automatically. The two options were to either edit the tape to remove the unwanted takes, or to visibly mark it up to make identifying them easy, which also had the advantage that alternative takes remained available if needed and could still be dropped into sync in the same way.
Early on when doing dubbing in the Sypher suites, those who had taught me the job, especially Ian Leiper, drummed into me how important it was to properly mark up the script, but when I did complicated sequences, I found it tricky to keep tabs on what was on each audio track and to see what’s coming up. Part of the problem was that my script was already marked up with notes made during the shooting and there was often a lack of empty space. My solution was to use my home computer dot matrix printer to print vertical lines with two or three minutes per page corresponding to the audio tracks and grams feeds. It could be aligned with the faders and had minute and seconds markings along the vertical axis. Being fan-fold paper, it was continuous and could be printed to be either 30, 60 or 90 minutes per episode as required. The printout included stuff I already had planned prior to the dub and would additionally mark it up by hand as the dub progressed. During the dub it was easy to see what was about to happen on each fader and felt very natural. I was rather proud of this system and some years later when I was being shown around the Ealing Studios by Roger Turner, I noticed their dubbing sheets, with vertical coloured-coded columns representing what had been laid on what track. I felt somewhat miffed that they had copied my idea about twenty of thirty years before I ever thought of it. Digital dubbing systems now use a horizontal scroll, but in the 90’s, a digital system called Dawn also incorporated an option for vertical scrolling which I liked using, but few people ever seemed to use it and it was eventually abandoned.
One of the things I hated about the BBC was a tendency for people to work within ivory towers, I was always curious about what other people did and how things work, so it was natural for me to see what could be adapted from other disciplines to make my work better. I was also very fortunate to start my supervising career with line managers such as Colin White and Jeff Baker, who both encouraged such experiments … or maybe just humoured me so that they could have a quiet life.
Tony Nuttall
A couple of superb explanations of Operational Techniques by Alan Taylor and how to achieve them. So interesting to see Alan’s thoughts on sound problems and his mental provision of a solution! The other plus was his way of thinking that provided a solution with existing kit. A fine example of a BBC member of staff who encompassed all of his BBC training and production experience etc etc to provide exemplary sound.
++++++
Before moving into the movie industry – and working on feature films as described above – Pat Heigham was a BBC TV Technical Operator. Here he recounts how grams ops worked on the iconic BBC program “Doctor Who”.
Pat Heigham – Dr. Who’s Doing What?
This photo appeared some 53 years after it was taken:
The photo was sent to me by the chaps who are restoring the very first episodes of “Dr. Who”, believed to be from film telerecordings.
It was in the property of the late Verity Lambert, who was the executive producer of the early “Dr. Who” series and depicts Adrian Bishop-Laggett and myself, with Verity, in the Sound Control room of Studio D, Lime Grove (sadly non-existent anymore, flats now stand on the site).
Bish and I think it must have been around 1963-4, as he was the Grams Operator at the time and I was double-banking him, prior to taking over that job on his promotion to Sound Supervisor. I stayed with the show for about three years, had a break and came back to it again in 1968.
As a rule, the usual routine for working on the series, was attendance at ‘outside rehearsals’ in a drill-hall somewhere, to witness a run-through, a chat with the director if anything special was required in the way of sound FX, then back to Television Centre for a pie and pint in the club, before hitting the Gram Library to audition FX recordings needed. After that, a short walk down to Studio R in Lime Grove, a dedicated sound facility, to assemble the reels of tape to be used at the studio recording, which was done in one day for a half-hour episode. The programme was videotaped to 2” Ampex (405 line B/W in those days!) and any recording breaks meant that the videotape had to be physically cut and spliced. Spot FX were played in at the time, and any bridging music links or changes in background atmosphere were added over the edits in a re-recording session later.
Many of the sound effects were created by the Radiophonic Workshop at Maida Vale Studios and I often drove over to liaise with Brian Hodgson and collect any specialised material. I do remember belting my mother’s twin-tub AEG washing machine which produced the most wonderful array of motor whines, switch clonks and whooshes, when I could take home a Uher or Nagra tape recorder.
The tape machines in the picture are EMI TR90s – a beautiful piece of kit, running at 15 or 7.5 ips with an amazingly fast start, which made ‘spot’ cueing a reality. They were two-track, which was useful for long atmospheres, as the switch between them was done with a quick flick of the faders.
BBC rotary faders had a 180° travel only, so from zero to flat out was extremely fast. Later programmes, using the sound desks in Television Centre (TVC), had the facility of a sound source auto-following the vision mixer cuts.
It is possible to see, disappearing off the nearest machine, a long tape loop, which extended across the gallery and round a cine spool mounted on a pencil taped to a mike stand! It was a challenge to edit the loop so that the audio join was not discernable on either of the two tracks. No computers in those days!
The Dalek voices were largely down to the actor Peter Hawkins, who produced most of the characteristic metallic rendition, helped a bit with a ring-modulator which put in the ‘broken-up’ pattern. If there were multiple Dalek speeches, then some were pre-recorded and played in off-tape.
I remember one occasion when I looked down at my script, realised that I had not cued the last dialogue line of the scene, and left Deborah Watling (Victoria) with egg on her face, waiting to react!
As most sound FX were relayed to the studio floor (foldback) for actors to react, I did have a bit of fun with Carole Ann Ford, who was due to leave the show, as ‘she had fallen for a chap on another planet’ and was not going to travel with grandfather anymore. Her dialogue line: “Oh, David, I love you, I really do!” was absolutely asking for the addition of the soupiest, schmaltziest Hollywood strings that I could find – and she got them! It stopped the rehearsal, with her saying: “Well, I’ve never been sent up like that before!”
Endnote:
I was recently asked to join in to a ‘bonus’ extra commentary item for the DVD release of “The Ice Warriors”. It was good to link up again with Deborah Watling and Frazer Hines.
I related a story involving Deborah. Technicians were discouraged from fraternising with artistes – BBC protocol rules! On one occasion, I was having a snack in the Club after a “Dr. Who” rehearsal and Deborah came in with her father, Jack. They asked to share my table, whereupon, she proceeded to play ‘footsie’ with me under the table! Very sadly Debbie passed away in July 2017, from lung cancer.
(picture shows from left: Toby Hadoke (interviewer), Frazer Hines, Debbie Watling, self, Sonny Caldinez (a very large Ice Warrior).
© Pat Heigham 2017
