Theatre soundWhile centred on the stage, preparations for a major theatre production put sound operators on an equally steep, if not steeper, learning curve than that of its actors. Not only do they need to learn every actor’s lines and cues, but they need all of the show’s effects in their head and at their fingertips.

Andrew Bruce lifts the curtain on the action at front of house, from first rehearsal to the premier and beyond.

On a big budget musical, the period from the first day of rehearsals to the opening night can be as long as 12 weeks. It usually follows a set pattern with a recognisable series of landmarks – but, because it can be such a slow burn punctuated by long periods of apparent inactivity, it requires a certain degree of stamina and quite a bit of willpower to avoid the inclination to put things off.

ScriptAbsolutely everyone connected to the show will meet on the first day to introduce themselves, but probably won’t all come together again for another six weeks at the earliest. After that day (when everyone’s on best behaviour), we’ll each go into our respective rehearsal modes – actors learning lines, musical folk teaching songs and harmonies and the director and choreographer blocking scenes and steps while scenery, sound and lights start to load into the theatre.

During this time the sound designer might only put in an odd appearance at rehearsals, while the operator, as his representative, will watch, listen and learn but above all, become an actively participating member of the company – this last being very important at an early stage.

It all starts off very slowly. He/she hasn’t got an awful lot to do except gather snippets of information from the director, choreographer, costume designer, scenic designer and even the script writer, while noting the need for any sound effects that aren’t already obvious and watching out for blocking that may cause possible future problems. So the operator can get on with learning the script, getting the measure of how the scenes mesh with each other, and listening to how the principals deliver their lines – as a good operator and a singer will develop in a truly symbiotic relationship as time progresses. This phase also gives an opportunity to gather information from the musical supervisor about the distribution of parts amongst the singers – which of them have the strongest voices, which ones to avoid and which ones to leave out of the programming entirely because they are actually dancers who only sing a bit, when they get their breath.

Every scrap of information gathered here will become really important to the mixer later on.

Meanwhile, the hardware is going up in the theatre, but for the operator there’s relatively little to put onto the console apart from translating the sound designer’s plan, naming channels and experimenting with the layout to make things logical and easy to find. As yet, there is no session apart from a cue that sets the console into an initial state – it’s all about lists and script work, deciding on points in the script where cues are needed to assign channels to the control group surface. But as rehearsals progress, cues begin to materialise thick and fast, with the operator building cues in offline software on a laptop and daily delivering an updated session that represents the progress to the theatre after rehearsals have finished.

Stage call

Mother Courage

Once the show is blocked, work in the rehearsal room finishes and the acting company decamps to the theatre. The operator will, by now, have built up a basic session, with cue points to change assignments and trigger both processing and sound effects. But when the actors come on stage, they’re usually wearing neither costumes nor hats but they are wearing radio mics for the first time although at this stage they’re often less focussed on actually singing. For them, understandably, what’s important is where they stand on stage, which way they’re facing and where they enter and exit.

For the moment, singing full out is the least of their worries because they’ve been doing that in rehearsals for the previous six weeks, so the operator is faced with a load of mics, often roughly placed, on a group of people – many of which are only half singing. After the air of anticipation that precedes the arrival of the actors, it can be singularly demoralising for the sound designer and operator alike – for a while.

Mics on heads are a real compromise. We hide these tiny omni mics all over people’s heads, ideally the forehead, or wherever we can get the best sound without burdening the actor with an obtrusive wart, which is not always avoidable. Ultimately success or failure depends entirely on the shape of the performer’s face and their hair – or lack of it – in which to hide the cable. Actors who shave their heads only have themselves to blame! Much of the time the mics are not really in the ideal place – either because there’s just nowhere better or they’ve slipped in the heat of battle.

Wizard of Oz

It’s almost the worst possible scenario if you want to get a good vocal sound, but it’s the best we can do and at least singers don’t go off mic when they turn their heads – as they used to when the mic was on the body. So, initially, the operator is trying to get to grips with the system but with very variable input. Although that aspect sorts itself out as the actors become more comfortable, all we get for the next three weeks to represent the entire orchestra is a jangly old piano playing somewhere in the stalls or the orchestra pit.

Nevertheless, this phase gives the mixer an opportunity to really learn his script because, what is not generally understood by people outside theatre, is that while we follow every single line, at times, there may be anywhere between 35 to 50 people milling about in close proximity on a 15m by 10m stage, each one wearing an omnidirectional mic and all facing in different directions. Added to that, the PA may be only a couple of metres away and with foldback contamination everywhere. You cannot afford to leave any unused mics up at all – ever – so you’re not just following every line, you are killing every unused mic stone dead, even between lines. The mixer has to know not only everybody’s script, but also their delivery, timing and inflections as well as their vocal strengths and weaknesses.

With potentially so many omnidirectional mics on the stage and on the move, the avoidance of phasing (comb filtering) is an art form in itself and, frankly, not always possible. You often get two singers facing each other, a metre apart or less if they are in an embrace, singing or harmonising while swaying ever-so-slightly but significantly when it comes to electrical summing. If the operator fails to clock what’s going on, you will get the most awful phasing so there are a number of techniques and devices that are employed in order to counter that.

The simplest technique is just dipping one mic, because as long as the two are close enough together and of similar heights you may get some semblance of a phasing-free balance that way, but only as long as you dip the mic on the singer with the stronger voice. Of course that might end up being for only a few words, because as they move or turn away you’ve got to be ready to bring it back. If the height differential between the two is too great (the men are usually taller) and the two are very close, the technique won’t work at all, because the man will be singing directly into the girl’s mic and she won’t get a look in. In which case, all you can do is to appeal to the director to change their blocking (very unlikely if it’s a love song) or at least cheat their positioning so that from the front, they appear to be closer than they really are (really hard to do under stage lighting).

It goes without saying that all of this is beyond the scope of any automation because they may not turn on the same word so the operator has to be constantly hands-on and watching like a hawk.

Cats in hats 

By now we still have the piano, and we have a set of actors who are comfortable on stage and once again concentrating on their singing. We might have some costumes but they generally don’t make much difference except that you now have somewhere to hide the mic packs. But what does make a huge difference is when they introduce the hats. A microphone on somebody’s forehead in relatively free air has a certain, not unpleasant, sound but add a hard-brimmed hat half an inch away from it and reflections from the brim instantly introduce a dreadful change to that sound. Up to the point at which digital consoles arrived in the theatre, it was almost impossible, while mixing line-by-line, to reliably find the time to make any meaningful EQ changes to compensate for the unwelcome effects introduced by hats.

HatsOnce again, it’s something that the operator has to watch intently – when they put the hat on, take it off, put it back on again – so the ability of a digital console to provide us with some help here is very welcome. In the days of analogue, finding the time to make the EQ changes to one channel was hard enough, let alone two or three that might require it simultaneously. Often it was one of those things that was brushed under the carpet – there were some horrible-sounding vocals when hats were used in those days.

It became obvious to us that one of the most important things that a digital console could do for the theatre was to allow us to define a rolling, (updatable) subset of EQ settings that was associated with any particular character’s hat and that we could drop into the cue-list whenever the actor put it on. We christened it an Alias, as that seemed to admirably describe what was happening.

Hats are rarely used just once or twice, they often reappear throughout a show, but often you only get a few words to experiment with before the rehearsal stops or the hat comes off again, – very short bursts in which to hear and analyse the effect that the hat produces. The concept of the software had to be that you could repeatedly return to the same Alias throughout the cue list and continue from where you had left off. Any corrections that you had previously made when the hat first appeared would be automatically available wherever you chose to use it again.

In short, with Aliases we can continuously refine any hat EQ irrespective of where we are in the cue list and never lose anything, secure in the knowledge that we’re not upsetting any other EQ settings associated with that actor. It’s all a work in progress.

Orchestral manoeuvring

Eventually we get to the day when the orchestra arrives. The load on the console and the operator rises a hundred-fold, not least because everybody needs foldback. There are unique and often contradictory problems that occur when you put large orchestras in small Victorian pits. Trying to keep everyone happy and limit the amount of noise clutter in a pit that’s too small is an art, but not a very rewarding one. Often, the pit is so cramped that players are far too close to each other and the mics are far too close to the instruments – and yet the layout of the pit is such that sections that need to play together cannot reliably hear each other so you must resort to giving them foldback. At the same time, not only do instruments (and foldback) bleed into each other’s mics but players frequently find it hard to play and hear themselves because of the proximity of other players. And, if you manage to find some room for screens, they no longer play as an orchestra because they can’t hear each other naturally, so you have to rearrange the layout or… put in more foldback.

Andrew BruceAt this point the mixer is still learning lines, we’re refining the baseline EQ on the actors; we’re learning all about new hats that are arriving daily while dropping hat aliases into new cues to take care of the temporarily altered sound of the voices. As well as that, now you’ve got maybe two or three other people making adjustments to another 60 or more orchestra channels all requiring EQ and foldback simultaneously, while the sound designer is trying to hold it all together. On the day the orchestra arrives, we send the actors away so that we can get a nice balance up on the PA and for the first time things start to sound good. But you realise that there’s very little point in doing that for long (except to make the composer happy) because soon you’re going to have to put voices against that orchestra and everything is going to have to take its rightful place in the auditorium mix which often, to a greater or lesser extent, takes as its reference little more than the level of a subtly reinforced live voice.

At nine weeks, you’ve got the orchestra and the singers and the sound effects, and it’s all finally coming together. You’re going to get another week or so with the orchestra but not every day, because it’s far too expensive to have them sitting around while there are still so many staging elements to fix. You get them for a day then just as soon as everyone, particularly the actors, is beginning to feel inspired and a bit comfortable, you lose them again and return to the piano, then you get them for another session, then back to piano. You stagger, slowly and disjointedly, towards the first preview…

There will always be a couple of dress rehearsals, at least one, before the first preview, then a clutch of previews. Nowadays it could be 30.

Since this will be the first time the show has truly run to speed and in front of an audience, there will inevitably be elements that don’t work or the audience don’t get, so there will be music rewrites, script rewrites, whole-scene rewrites – and the mixer will have to react to all of them, sometimes dropped on him 30 minutes before a show. Sometimes whole scenes are cut, only to be put back the following night, so you need to be able to skip over cues in the cue list without removing them unless you know they are definitely gone. You have to be sure that the programming flows through, even though you’re skipping cues.

Usually, about three days before the show opens it is ‘fixed’, so that everybody gets to do a number of shows without any more changes and can settle into their roles. After opening night there’s a big sigh of relief because you think nothing is going to change any more – but you couldn’t be more wrong. The alternate players and understudies start to kick in and all the EQ that was done for the principals may have to be temporarily shelved…

See also:

Andrew Bruce: The Musical
(The development of DiGiCo's theatre mixing software)
(The evolution of mixing sound for theatre) 
Theatrical Address: Loudspeakers for Theatre
(The history and development of theatre loudspeaker systems) 

Pioneering theatre sound designer Andrew Bruce is the founder and owner of Autograph. He began his career at the Glyndebourne Festival Opera before becoming Head of Sound at the Royal Opera House, Covent Garden. His extensive musical credits include Evita, Cats, Les Misérables and Miss Saigon.

TwitterGoogle BookmarksRedditLinkedIn Pin It
Fast-and-Wide.com An independent news site and blog for professional audio and related businesses, Fast-and-Wide.com provides a platform for discussion and information exchange in one of the world's fastest-moving technology-based industries.
Fast Touch:
Author: This email address is being protected from spambots. You need JavaScript enabled to view it.

 
Fast Thinking:Marketing:  This email address is being protected from spambots. You need JavaScript enabled to view it.
Web: Latitude Hosting