The following is a detailed breakdown of all the sound design elements in a show from last season. I hope that by touching on all the elements, it will help the young designer understand the context and relative weighting of the different responsibilities of the designer.
At this particular house, I mix my own shows. This creates a small amount of conflict between the different roles of FOH (Front Of House mixer) and Sound Designer, particularly during the height of Tech Week. In most shows I've had someone to "push the button" to run sound effects, but I've rarely had a skilled A2 (Audio Assistant) who I could task with anything more complex.
The building seats 350. It is a converted church, rectangular in shape with a gently sloping seating area and a high raised stage. The installed sound system is a Meyer design (and contractor install), consisting of a pair of mains flanking the stage, a second pair of speakers 3/5 of the way towards the back of the house, subwoofers (currently in the orchestra pit) and a pair of front fills hung high over the front edge of the stage. There is a Galileo processor running all these speakers (corrective EQ, delay, and relative mix) and that in turn is fed from the FOH desk; a Yamaha LS9 digital board.
It is a good system for the space. If all we were doing is ballet and live music, we could just send a stereo pair back to the Galileo and have it locked into a carefully tuned preset. As it is, we are largely doing musical theater -- which places a combination of reinforced and natural voices against a live band.
Two other peculiarities with the kinds of musicals we do. We tend towards a large mixed-age cast, meaning there are always choruses that are not on microphone. And we tend to place the band on stage with the action. Both of these help make getting the vocal reinforcement right, and getting the monitor system right, challenging. And as a result I deprecate ordinary sound effects -- I do them, and I do elaborate ones, but they have to take second priority.
To overview: The responsibility of the Sound Designer is all sound elements of a production. This is usually broken down into;
Vocal Reinforcement (for live musicals, this is mostly wireless microphones).
Orchestral Reinforcement (for some shows, replace this with the cuing of pre-recorded backing tracks).
Performer Monitors (audio and visual links between musicians and singers).
Backstage Monitors and Communications (intercom system, Green Room monitors).
In many houses, the Green Room monitor, and the headset system for crew, will already be in the building and you won't have to worry about it. But you can't count on this!
It is also your concern when there are sources of noise on stage -- squeaky platforms, noisy air conditioning, street noise through an open door. You won't often have the authority to do anything about these, however!
This particular show under discussion is a (relatively) new musical by the name of "Lucky Duck." I'll discuss sound effects first; they may be lower in priority than getting the vocals right, but they are fun.
"Lucky Duck" joins such shows as "How To Succeed in Business" and "The Will Roger's Follies" by having a pre-recorded performance that is central to the action. In "Lucky Duck" the NARRATOR interacts with the cast. In such cases, getting those VO (voice-over) sessions in the can is utmost priority -- you want them in rehearsal, weeks before the show even goes into Tech, so the actors can work with them.
Fortunately, in this case the NARRATOR was played by a cast member. So he did the voice overs live through rehearsals, and when we felt the performance had matured enough, we recorded them.
So the very first sound effect was a VO recording session. I've talked about these before. You want a space with good sound quality to record in. Not so much quiet, as one that lacks distracting reflections that are difficult to clean from the audio. Extraneous noise is easier to remove than the constant reverb tail of too live a room. And with a strong vocal performer, even a fair amount of leakage of general noise can usually be compensated for. So...look for a room with soft walls. Or a big room. Failing the above, find a cluttered room; the clutter will break up the reflections and make for a softer, more controlled room tone.
Make the actor comfortable and have time to work. Have the director present -- both for their artistic input, and for their skills in getting a good performance out of an actor. Print out the material to be recorded, with large type and generous margins. There are a few other tricks...like green apple slices, which I've never used... but you get the drift.
In this case, what the actor had been doing in rehearsal was interesting, but didn't work as a recording. It was very shouty -- a strong projecting voice for someone addressing a crowd, but the wrong feel for an omniscient story-teller looking over the action and commenting on it. So we worked at getting him to speak softly but with power. The final mic position was about 6", a large-diaphragm condenser suspended just above his forehead looking down.
When we finished the sessions I had almost three complete takes of all the material, plus a lot of false starts. When I went into cutting takes, I ended up making hybrid cues for two of the more critical ones; cutting in phrases from two or more different performances. I did a very small amount of slicing in taking out extra pauses, deleting some breath noises (but not all; I needed that human presence), and of course editing out several plosives. Then hand-normalized, EQ, and gentle compression.
The same actor also read a different pre-recorded vocal, the "Quakerdome" announcement. Which I gave the "Monster Trucks" treatment with delay, heavy EQ, and sub-base synthesizer ("Sunday! Sunday! Sunday!")
For this show I didn't do any Foley. Well; I call it Foley, but technically the process of Foley Walking is the real-time performance of small noises (feet and clothing mostly) to a chunk of film. What I do is mostly recording bit and pieces of me fumbling around with various props...walking, dropping boxes, playing with leaves, etc., which I then process to make effects. In this show there were, as far as I remember, none of these.
The other effects came mostly from library sounds, plus new sounds purchased from Sound Dogs dot com. I wanted a European two-tone siren for a different feel to the urban scenes, for instance.
The sounds that were the most fun were the chop-socky effects. When DRAKE comes to the rescue late in the second act the actor improvised -- hilariously -- a bunch of wacky-looking martial arts moves. To make it even more ridiculous I pulled a bunch of punches and kicks and swooshes and assigned them to a MIDI keyboard -- and did my best to Foley his live performance every night.
Usually my best was not enough, but it worked well enough to be funny. The effects themselves? I ended up going to YouTube, where I found some collections of the proper badly-dubbed 1970's Hong Kong import type hits, chopped them and cleaned them and so on. I also assembled a long swoosh for when he comes in on a rope; this included a swoosh from a YouTube video of a Foley session from "Kung Fu Panda" and a fluttering sound from a small sailboat (courtesy of the BBC sound effects library).
I find sound effects fall (though not always neatly) into three general categories;
Ambiance (sounds of the background environment, which play under the action)
Spot Effects (which happen at a specific moment in the action, such as a gunshot)
Transitional Effects (which happen between scenes)
For this show, there were two or three ambiances. The "Scary Forest," the "New Duck City" urban space (trolley cars, traffic, distant sirens), and the "Sewers" (various drip sounds and some steam and a little water trickle, all with delay effects).
The forest opened up, for me, the Goofy Question. This is a basic problem for any setting that includes anthropomorphic animals (in "Lucky Duck," ducks, wolves, coyotes, and at least one armadillo). Goofy has a dog, a dog named Pluto. Pluto walks on all fours, has a leash, and doesn't talk. So what is Goofy?! Maybe the best general answer is to assume settings like this, like Narnia, have animals and Animals. And that explains why a duck can eat fois gras (shudder!)
I'd previously done "The Jungle Book," and "Seussical," and followed this show with "Honk!" and "Click Clack Moo," and I determined early on that no "animal" in the show would make a non-human sound. I'd mixed "Narnia," earlier -- different designer -- and he also made that choice. In that particular theater, we don't hide the theatrical nature of the performance and whenever possible we have the performers do what needs to be done. And that includes making animal noises.
But that meant I really wasn't free to have howling wolves or lion snarls or anything like that in the Scary Forest. And rustling leaves just wasn't enough. And I couldn't even use birdsong, because most of my cast were birds! So in the end I added some insects and snakes. Because as far as I could tell, neither of those kingdoms were represented by actual cast members in the show.
There were a couple dozen spot effects but I'm only going to mention a few. There was the wolf trap being opened (a snippet from a rusty gate sound). The "Sorry!" buzzer for the talent show (which was a stock buzzer sound re-pitched and shortened). Actually...the buzzer is what made me stick an XBee into a Staples "Easy Button," and by the end of the run it could have been a practical sound effect (as in, the actor would be triggering it from stage). We stayed with making it a traditional, visual cue however. The zap sound for the invisible fence was a combination of two different "zap" sounds, one short and one longer snippet from a Jacob's Ladder effect.
I tend to program ambiances so they start loud to establish, then fade down. Some of these fades were called cues. Some were built into the sound effect. Most were programmed into auto-follow cues in QLab. Once the timing had been tweaked during tech rehearsal, they played themselves automatically.
For this show, one laptop up in the booth ran all the "Called" cues. Actually, the Stage Manager was pressing the button himself so he didn't actually have to "call" them. A second laptop was connected to a MIDI keyboard and from there I ran improvised sounds such as the Kung Fu stuff and some camera and flash camera stuff for the big fashion show.
An alternate arrangement I've done in the past is to have at least four channels routed from the primary sound effects computer; one pair goes to spot effects, the other goes to ambiances. This gives me the ability to mix in the ambiance effects to the total sound picture from the FOH position without messing up the volume of the spot effects. Another trick I've done is have pre-recorded vocals show up on the main vocal bus so they can me mixed and placed together with the live vocals.
In this particular show the sound effects were simple enough, and the way they fell was developed enough (aka I didn't have to wait for set pieces to be delivered or scenes to be blocked before creating them) that they were essentially delivered by early on in Tech Week and I could concentrate on other responsibilities.
Sound effects are exquisitely sensitive to the actual playback environment. What the orchestra is doing, what kinds of noise the scenery is making, the timing of the action on the actual delivered set or with the actual production prop, and of course the specifics of speaker placement and system tuning and the acoustic environment.
I bring the laptop with all the material on it to the building. For this show I had the time to set up my laptop in the middle of the house and playback sounds from there as I adjusted them to the space. The Scary Forest cue took more time than that, because the sounds of the actual scene worked differently than the sounds of the stage in an early afternoon with the lighting people doing a focus. So I had to take it back and adjust it several times.
I was never completely happy with many of the cues, but they worked well enough. And with it loaded on to a laptop, I could move on.
This was one of the most easy-going pits I'd worked with in this space. Many music directors are, I am convinced, half-deaf and they usually demand stomping monitor levels to be able to hear the vocals over the keyboard amp that they also turn up to ear-splitting levels.
Volume wars in most spaces start with the drums. It takes a good drum player to play softly, and very few of even the good players like to play softly. So the drums are too loud already, and they get into the vocal mics, and they make the audience complain. But that isn't the worst. Since the drums are so loud, the bass player has to turn up his cab in order to hear himself play. And whereas a nice focused DI sound can be tight and musical throughout the room, the almost completely omnidirectional sound of a bass cab turned to eleven saturates and permeates the space with a mushy, undefined low-end rumble that completely destroys any hope of getting an intelligible mix out of it.
And with bass and drums committing murder in his ears, the keyboard player has to crank up his own monitor amps in order that HE can hear himself. And when he is finally comfortable, and the bass player is happy, the drummer suddenly realizes he can hear the other guys again and he doesn't have to lay back in order to figure out where he is in the music. So he gets louder. And the wars go around again.
As hearing fatigue sets in, each player needs more and more volume just to maintain bare audibility. And their monitors -- with an unfocused, bassy, echoing sound -- are washing over the entire audience with sonic mush. And as the poor FOH, the only real choice you have is to crank the reinforcement to scary levels just to put some definition back into the sound.
And that's before you add the vocals!
For this show, fortunately, I had a relaxed keyboard player who could handle being able to hear less than he'd like. And a bass player who was happy enough just to use a monitor we supplied, set up on a chair right in front of his face, and not blast everyone with his own cab. And a drummer...well, one of our drummers was more restrained than the other. But we basically made it through anyhow.
The most important element in the pit is a send from the vocal bus to a monitor -- I use a tiny "Jolly 5a" from FBT set up on a mic stand to get it as close as possible -- that the conductor can use to hear the people on stage. Second after that is a send from the keyboard -- and sometimes bass and kick -- to front fill and side fill monitors for the actors. That way, the actors on stage can hear equal levels of the music they are trying to sing to, regardless of whether they stand next to the band or in the corners of the stage.
My choice for years has been low-profile monitor wedges placed along the front edge of the stage. I use my FBTs a lot because of their extra-wide pattern. I also fill from the wings; generally behind the second set of wings pointed in and slightly upstage.
When the monitors on the stage become so loud they are a major part of what the audience hears, I send monitors a full band mix instead of just keyboards. This way, changing the monitor level doesn't unbalance the orchestra mix.
One issue with orchestral reinforcement in a smaller space like this is you don't have a second mixer to deal with the orchestra. At the FOH you are already dealing with 16-24 channels of wireless microphone, plus a few extras, and trying to actively mix the band is not really an option. Also, in the theater I have been describing, our LS9-32 has no add-on cards or digital snake and those 32 input channels get burned up quick between wireless mics, sound effect playback, and extras like talk-back microphone.
The reality I've found in most shows there is I am pushing to spare six input channels for the pit. And in addition, I make a lot of use of the custom fader layer in order to have fingertip control over mix buses, so those six channels of band have to be thrown into another layer -- which means they aren't at my fingertips during the show.
On several previous shows I've used a second mixer to create a band submix. The designer for "Annie" provided me with a complete 24-channel Brit board with a rack-mount compressor and effects box to go with it. I've used my own much smaller Mackie to the same effect; on one show I added two more mic-level channels to the four already there via an "Audio buddy" pre-amp and used that to create a drum sub-mix.
For the punk-rock band we had in one show I went all out and put mics on kick, snare, hat, cymbals, and low tom, and I submixed those on a laptop computer running CuBase. And every night I was terrified the computer would crash on me in the middle of the show.
These days I'm tending towards kick and overhead. With a four-piece band -- such as for "Lucky Duck" -- it knocks down to keyboard (DI or direct out from keyboard amp), bass, kick and overhead, and wind. This particular show was a multi-wind player. I set up an overhead and a low sax/clarinet mic, but discovered during the run the overhead was the better sound for everything. I flipped that fader to the top layer of the LS9 because I was riding it a lot during the show. The rest of the band, however, I could leave alone.
This means the default in that building is 6 inputs from the band. I run a 12x snake to them, and I dedicate 1-8 for inputs and 9-12 for monitors. 9 is the "Con-Mon"; vocals for the conductor's monitor. 10 is general monitor for the rest of the band with 11 and 12 for anyone that really needs their own mix.
Except that more often than not, I'll stick a mixer down in the pit with them and, by using the AUXs, the conductor can adjust the mixes sent to two different players while also maintaining his own mix of keyboard, vocals, and any band elements he needs to hear more of (usually none!)
The big downside to this method is that when volume wars start, the conductor can crank their own vocal monitor as much as they like...ignorant of the very real feedback limits the mix sent to the house is up against. I've had the conductor's monitor end up so high IT starts feeding back.
Bass often ends up on a mic instead of a DI, not because it sounds better, but because I've had so many bass players in the past with hideous ground loop problems. They also have a habit of unplugging 1.5 seconds after the last note, and I can't always get to the mute in time!
I've really fallen out of love of close-micing snare, which has never quite sounded right even with heavy gating, compression, and corrective EQ, plus a bit of room reverb. The best snare sound I ever got in that house came from a saxophone mic; four feet away from the snare and pointed 180 degrees away from it! For kick, I've been very happy with my $60 Gear One, stuck just slightly inside the vent when they have one. You get more of a papery sound if there is no vent in the front head. I've tried using a short boom stand to get a mic up tight to the beater head, but musical pits are cramped and hard to access and without an A2 you are at the mercy of wherever the mics got kicked to as the band wriggled into the pit four minutes before the downbeat.
Winds and strings speak best with some air. The most natural sound is a couple of feet away. In a cramped pit, though, you get too much bleed (especially drums), so you end up compromising. Also, winds are always multi-instrumentalists in a pit orchestra, so you don't have the luxury of getting a close-mic in exactly the right spot on each different instrument. Heck -- often they are switching fast enough to chip a tooth (real story!) So set up a good condenser around six inches to a foot, depending on how much that particular player jumps around.
And those are the basics. As few mics as possible, set so they don't have to be tightly positioned every night but have some slop in them, and sub-mix them so you aren't staring at those faders in addition to all the wireless.
(Actually, I think I confused some of this with the pit for "Honk," which shared the set. The pit for "Lucky Duck" was a bit louder, and sometimes I had to fight to get the vocals up over them. But they were still a pleasure to work with.)
The theater owns 18 channels of Sennheiser "Evolutions." After that, it is a few units of Sennie G2's, and some Shure SLX we mostly keep around as hand-helds. Lately I've been loaning some of my own seven channels of Sennie to bring us up to 20-22 working channels on a typical show.
"Lucky Duck" actually has relatively few named roles, and even fewer players (there's a fair amount of doubling.) There are only a few chorus numbers. At this company, we almost always include a younger cast; they are treated as a separate chorus and brought in for specific scenes. They are almost never on mic, however, one of them did get a mic for a few lines of dialog in this show.
Unlike many shows, there weren't any major issues in finding places to hide the mic packs, and allow the actor to do the various physical things we often ask of them. The worst case was probably SERENA's Act II dress, the scooping back of which showed off the dangling mic cord to good advantage (there's only so much surgical tape an actress can put up with).
My default mic position is forehead. Because of animal costumes and hats we had to move to cheek for a number of the cast, including essentially all of the male cast. Also, our SERENA had the pipes but to get that intimate pop sound we chose a B6; with the flexible wire boom that allows it to be placed close to the lips, Madonna style.
Because these songs are very mic'd and very pop we were scared of losing a mic on stage. So we gave the talk-back mic to the drummer so he could hand it up in case of a body mic failure! We never actually executed this safety but it was fun knowing we had it.
In many shows I've had microphones set up for an off-stage chorus. Not this one. I've also sometimes had floor mics or spot mics for an un-mic'd chorus. Again, not this one; this was a very simple show wireless-wise.
The Meyer install is very nice, but there is a major flaw when dealing with a live band. And that is taper. Since instruments like drums are acoustic, they are naturally loud for the front rows, and softer for the back rows. If you were to leave the system settings flat, with the rear delays covering the back part of the house, the keyboard would be softer than the drums for people in the front rows, and louder than the drums for people in the back rows.
So I've set up custom programs for a slight taper to the system. Also, I've chosen not to send vocals to the subwoofers. And lastly, I use some of the house system for sound effects playback -- which means I need access to specific speakers outside of the general programming.
The optimal solution in that building turns out to be the non-simple one. There are six inputs to the Galileo speaker processor, which in turn is feeding 8-10 speakers (depending on whether we use the current center cluster, which was re-purposed from where they were in the original install).
So what gets snaked out to the Galileo are six of the omni outputs on the LS9. The Galileo splits them into several groups; right house, left house, sub, etc. In the case of the subs, for instance, the two boxes are being used in an end-fire or semi-cardiod array; one speaker flipped, inverted and delayed so the sound is boosted towards the front but experiences destructive interference to the back and sides, thus lowering the volume there. To get that effect I'm using two Galileo outputs and the internal processing, but I'm feeding it with only one omni from the LS9.
On the LS9 side, the mono bus and the stereo bus are each dialed to a different set of omnis, to give a slightly different sound; the mono bus functions as vocal bus and has no sub in it, for instance. However, to make it possible to send sound directly to the house system bypassing those two busses, mix channels 9-12 are patched in a one-to-one basis to the matrix channels driving the omni outs. That makes it possible for any input channel to behave like an independent mix bus and send to the house system itself.
(Well, actually, since several of the mix channels are taken up with monitor mixes, and several are in use for the internal effects racks, the special output buses are combinations of different useful Galileo inputs. And there is always a compromise between having access to a specific speaker for one specific sound effect, and being able to properly tailor the general house sound as fed from the sound effects computer).
Each speaker is delayed slightly to make the impulses arrive together at the audience. Each is also EQ'd a little to the room. But on top of that, the LS9 output channels are set for an overall delay of about 20 milliseconds. This makes the reinforced vocals arrive slightly after the impulse from the live actor, and via Precedence Effect focuses the apparent sound source on the actor instead of up in the flies.
On the vocal bus is strapped a graphic EQ for notching out the worst peak in the room (otherwise I try to run the bus, and the mics, without mucking up the EQ too much in search of every last db before feedback). I also often put a very slight bit of compression on the entire bus.
During the actual show, the mono bus is swapped onto the custom layer along with a linked fader for the sound effects, and the reverb outputs. With the band mix on the stereo bus, this puts masters for band, vocals, SFX, and reverb under the fingers of my right hand. My left hand is then free to nuance individual microphones.
Because this is a pop-sounding show, there are two reverb buses; a general "stage" algo to seat the voices in the sonic space (basically, to make them sound a little less like they are on mic) and a "plate" algo that gets cranked up for the more pop songs. Since I liked the longer tail on the "stage" for certain songs, I ran that up instead for those. In addition, mix bus 15 had a delay algorithm on it (again, from the LS9's internal DSP); this was applied to the vocal mics during dialog in the sewer scene.
The last effects bus is putting a little general reverb on some of the band. Usually, keyboard patches will send you a signal with reverb already on it. And bass sounds best dry. But I like a little on the drums, and I like a lot on certain winds.
On a different show the "special" effects bus (the one that isn't band, general vocal, or song reverb) was set to a ring modulator for a robot voice. I've also used the detune effect, but Yamaha DSP in that price range is pretty bad. For effects like OZ in "The Wizard of Oz" I use an outboard effects box. Unfortunately the LS9 has no channel inserts -- so this ties up a mix bus, an omni out, and another channel to do.
To follow all the mic ins and outs, we use the scene memory function. Each time there is a major change in which microphones need to be hot, I tap one of the User Assigned buttons, which calls up the next programmed Scene, which runs the motorized faders up. I tend to automate only the faders, using the Recall Safe function to lock out any other changes. This way, if I have a mic go bad during the show I can hit the Mute button and it will stay muted through following scenes.
Each scene number is recorded into a script, which I keep at the FOH and follow along in. After a lot of experimentation I've determined I don't like to try to hit every mic move. Instead I'll hit the big changes, and manually take out faders as needed during the scene. One trick is to program faders that need to be added in the middle of a scene so they come up to -40 or so; just enough so you can see it and know that fader will need to be brought up the rest of the way soon.
I try to leave things alone for much of the show. The big change is bringing up the vocal master about 5 db going into a song, and trimming it back down in following dialog. When there are duets, you often have to ride the faders a bit to get a blend -- especially when people start singing into each other's microphones.
The theater already has a Clear-com system set up; one channel, with the base station in the booth. In the past it sometimes got run through the sound snake but at the moment it is all dedicated wiring. Unfortunately, especially during tech we end up having to run long daisy-chains of headsets (because we have no splitters).
Over the years we've been moving out the old rack boxes and bringing in new belt packs. Also lighter headsets, and the first radio headsets (not very nice ones, though). My preference is not to use a headset at FOH, but just to have one available so I can call backstage and have a mic replaced or batteries changed. So that means that, barring failures during the run (when the usual staff isn't there and it falls to me to fix things) the headsets are not part of my responsibility.
Our current backstage monitor system, in common with several other theaters, is baby monitors. We've also sometimes hung an old dynamic mic in the rafters and run it back to a powered monitor. Lately -- and we did this for "Lucky Duck," I've taken one of the older wireless microphones we don't use and gaff-taped it to the set. They may be old, but the signal will still reach all the way back to the overflow dressing room. With an omni mic, it picks up the stage fairly well. All I need is, again, a powered monitor speaker (I have a pair of JBL Eon's kicking around I usually task with this) to connect to the receiver.
We do not have, at that theater, a backstage paging system, anything resembling a cue light system, or a wired communication from House Management to Stage Management. The place is small enough so the latter is not really a problem!
There really wasn't any "other" on this show. On other shows, in other buildings, I've had to set up the keyboards, even create sampler patches, set up electronic drums, run live microphones (for a production of "Annie" at another building we managed to score three period microphones for the radio scene -- so for that scene only we turned off their wireless mics and used the stand mics.) I even built/repaired cue light systems and headsets.
I'd say I spent under twenty hours creating the sound effects for this show, most of that being the NARRATOR (with twenty lines of dialog, it felt like prepping game audio!) There were no "Showcase" cues, like the B-Movie cue from "Grease," but it wasn't telephones and toilet flushes either; I got to have some fun creating effects.
Because this show was less fully programmed than many, and because I was doing a lot of tweaking of reverb settings, it got very busy during one or two numbers -- I was dialing in effects, riding the lead vocal mic, trimming the band, bringing in the chorus with a recalled scene and improvising camera clicks and flashes at a keyboard, all at once.
We did have a mic fail on a lead. Twice during the run we had to finish an act without a mic on one of the leads. On the other hand, I managed not to catch any back-stage chatter on a live mic. A connector got stepped on in the pit and we lost the monitor connection to the conductor in the middle of a performance. We also almost lost the keyboard but we discovered the bad connection just minutes before the curtain.
Quite a few body mics wore out during the show. We caught less than half of them in sound check; these things always wait until you are in the middle of a scene before they really act up. For most of them I was able to mute the mic and live without it until I could call backstage and get a replacement hung on the actor. By the end of the run, about 1/3 of the elements we'd started with had gone bad and had to be repaired. Which mostly meant a solder job, and usually a new connector.
And there you have it.