Saturday, January 28, 2012

Anatomy of a Cue

As a continuation of the discussion in the last post, I'm going to walk through the creation of a single cue in some detail. The cue in question is one of those lovely long, exposed, story-telling cues you rarely get to do; in this case, an audio slice of a drive-in movie. It is used in the production I just opened.

The voice-over session:

As with most cues involving dialog, the actual dialog to be used is specified in the original script (the show in this case is Grease and the lines in the B Movie sets up the drive-in scene and a song.)

In this case I didn't have to go looking for vocal talent; members of the cast had already been picked and had been doing the lines during rehearsal. The latter was a mixed blessing; although they didn't need to hold scripts, having already memorized the lines, that also meant I lost the chance to mark up the lines to better shape the line readings.

When I do voice-over work, I like to print the lines in big type, double spaced, one "take" to a page. The professionals will take the opportunity to mark breath pauses, special or problem pronunciation, emphasis needed, etc.

In this case I had flat, rote performances to start from. Working closely with the director we were able to delve into the meaning of the lines, find the important beats, and get those beats into the vocal performance. ("Beats" in the acting terminology sense.)

I've said this before; physicality is key. If I had time, I would have actually blocked the scene to give the change in voice that movement would cause. I was able to rework the second movie excerpt by requesting the voice actor playing the "hero" stand behind, with his hands on the shoulders of, the actress playing the "girl." This is a very, very typical couples pose in movies of the period. Looking over her shoulder like that caused the actor to give a more warm, comforting performance than the flat, affect-less performance he had been giving before that direction.

In a long-ago session, I recorded an actor seated, and had him rise from his chair when he reached a more emotionally intense motion. It can not be said enough; physicality shows up in the voice.

(The great, great story comes from the creation of the music for the Lord of the Rings computer games. The male chorus was giving a flat and lifeless rendition of the Dwarves song -- until the conductor had the men shift their weight from one foot to the other as they sung.)

Also unfortunately, we had only the theater lobby to work in, and it was raining. I knew I was going to get both room tone and extraneous noise on the track, but I felt I could probably work around it anyhow. Often in theater you have to accept what will work instead of what would be wonderful, because opening night is coming far too soon. And beside, if this cue is not as great as it could have been, you know there will be another show, and another opportunity, next month.

I set up an omni mic as back-up, but my primary mic was my home-made budget fish-pole boom.

I'm going to explain that, too. The fish-pole is the simplest kind of boom mic; nothing but a long stick with the mic at one end. The idea is to come down from above and stop just out of frame (that is, out of the frame of the camera for an actual movie). I've found it is also an excellent sound for voice-over work.

Putting microphones in front of talent causes many of them to deliver a performance to the mic; they get small, they talk into the mic. The voice you often want -- the voice I most certainly wanted for an imaginary scene from a B Movie -- is one that is large, space-filling, and directed out. So using the fishpole removes that obvious thing-to-talk-into and forces them to act to the room, to their partners, and to the imaginary audience.

A mic that is below the head, or even at mouth level, is a less pleasing sound than one that is aimed towards the forehead. This is why the hairline is the superior position for wireless microphones. A boom coming down from above and forward is a very natural, pleasing sound that mimics well how we perceive voices sound like in ordinary surroundings.

Here's the budget fish-pole (I should write another Instructable -- it was an Instructable I got the idea from!) Get one of those extending poles they use to change lightbulbs in big buildings. I found one for under thirty bucks at Orchard Supply Hardware. The fittings on the top screwed on with the same screw as found on industrial brooms and mops. I used a grinder to make the screw just a little smaller; until I could force a microphone clip over it. Then I mixed up some epoxy and stuck a universal microphone clip (another ten bucks) on to the end.

I don't have a Sennheiser MKH-416, but I do have a Shure PG-81; a mere cardiod (instead of short shotgun) but at the boom distances I work with it works just fine.

I boomed this time through my Mackie mixer, mostly for the headphone amp; this way I can wear headphones and hear what I am recording during the session. I followed the actors somewhat, shifting the boom a little to be closer to whoever was speaking at that moment. For such a short "scene," it was easy enough to memorize the necessary moves. Had they had blocking, that, too, would have been easy enough to memorize.

Of course blocking would have meant I would have had to walk around while holding up the boom...this is why actual good boom operators are valued members of the production sound team in the film world. I'm a pretend operator, totally self-taught, but I do it for the results I've heard in my voice-over recordings. (Plus, it looks cool and gives the actors a kick!) many takes as we had time for, made sure the file had saved properly to hard disk, and on to the next step.

Oh, and I knew I had a "mojo" take in the can. Most sessions, there will be one take that will make you sit up straight. Something about it cuts through the boredom of familiarity and makes the material fresh and exciting again. Nine times out of ten, that's the take you will end up using.

Processing the vocal tracks:

This has been a dry entry so far: let's enliven it with some pictures.

Here's the raw recording session in Audacity. I recorded at a basic 44.1/16 bit depth; the cue didn't call for anything more. In Audacity, I listened through the various takes and selected the take I would use -- yes, it was that "mojo" take -- copied that to a fresh file and normalized.

As I had feared, the rain came through. In an annoying fashion. If I had been more pressed for time I might have worked with the rain instead, but since it was a relatively constant sound I was able to remove most of it from the track with SoundSoap SE (purchased at half price as a special from Musician's Friend).

The trick to digital noise removal is to have a nice chunk of sound file that doesn't have anything on it you want to keep. The breaks between sessions, for instance. This is also another good reason to record a few minutes of room tone, without any dialog in it. After that it is a matter of ears and judgment to take out as much noise as is plausible without causing audible artifacts.

I read an interview with a production sound person recently and he stated the best way to do noise removal is to use several different methods. Every method leaves artifacts. If you turn any single process up until all the noise is gone, you inevitable turn up some objectionable sound. So instead you apply a bunch of different processes and as each leaves different footprints on the sound, those footprints are smaller and more easily hidden.

Anyhow -- SoundSoap was the first step, and that knocked the rain down until it wasn't objectionable. Now I could import the files into Cuebase and continue to knock them into shape.

Within Cubase I cloned the track, once for each speaker, then chopped each track until only the lines of one speaker appeared on it. This meant I could apply custom equalization and compression to each individual speaker despite them having originally been on a single mono track.

The gaps between their lines made this possible. But now that I was in an audio sequencer, I could also tighten that up a bit; I shifted several of the chunks of dialog in space in order to either close the gaps between speakers, or to allow insertion of an effect.

There was also a door that opened in the middle of the take. Of course this was the mojo take. Fortunately the door sound only occurred over one short chunk of dialog, so I pasted in those same lines from one of the other takes. The speaking tempo was different in that performance, however. But more luck; a time stretch operation, and not only did it fit the gap, it also gave the words more gravitas; it was a better line that way than what we had originally recorded.

I believe I may have applied a very slight pitch shift to one of the speakers as well, but for this project it was important to me to be honest to the voices of the original actors; to enhance them, not to hide or change them.

The girl's vocal levels shifted enough (my own clumsy boom operating was partly to blame!) and trying to fix that with compression would result in too funny a sound. Thus hand-drawn volume changes, akin to what we call "riding the fader" in live music, to bring it to a consistent level where the processors could work on it.

I worried at this point I might have to cut in room tone in every gap between dialog chunks, but I ended up going the other way: the lobby we recorded in was a little too "live" for what we were doing and I was getting some echo off the walls. Each vocal channel got, as a result, a hefty chunk of expander, set to an ultra-fast pick-up to close down the moment the last vowel sound left the actor's mouth.

Again this is a matter of listening carefully and balancing one unwanted artifact against another.

In period, dialog tended to be quite dry unless an unusual environment was being suggested (like an echoing cave). For that matter, there was a lot less production audio in the 50's; noisy cameras and so forth meant some films were entirely shot MOS and all the dialog picked up later in ADR.

Err...I'm showing off here with film terminology, and there aren't exact relationships to theater practice. "MOS" is filmed without sound. "ADR" is Automatic Dialog Replacement; the poor actor stands in front of a mic watching themselves on a screen and tries to lip-synch in the reverse direction (aka matching the words to the lips).

But this is also a philosophical question you hit every time you do a period show; how much do you want to be accurate to period, and how much do you bend to the expectations and perceptions of a modern audience? I have a byword I go to often; nothing is "old-fashioned" at the time. For someone living in the 50's, they were listening to top-notch, state-of-the-art studio sound. So we have a choice as a designer; to point a finger in laughter at the quaint past we are presenting, or to bring the audience back with us to experience an earlier time as the people back then lived it.

Anyhow...the choice made this time was to do relatively modern dialog recording methods. Or, to put it another way, dialog the way most of the audience are used to hearing it.

When I'm working on a voice-over taken on a close mic (say, for a radio announcer), I often have to manually edit out plosives. Another manual edit is when your actor manages to swallow a key consonant -- you can actually paste one in from a different part of the performance. But this is long, painstaking work and you really hope you don't have to get that detailed on your tracks (I had to do this once with quarter-inch tape and a razor blade, way way back on a production of Diary of Anne Frank!)


So now the dialog was done. The client apparently expected this is where my work would stop. I knew it wouldn't; without something to look at, raw dialog can be very, very dry and boring. I played the edited dialog track in rehearsal and it was obvious it needed something more.

The first thing I tried was filling some of the space with Foley.

Well, not really. In the film world, even when there is production sound the intent by the production recordist is to get clean dialog. Not all the other sounds. Film is a world of artficial focus. Instead of hearing all the sounds of an environment, you hear a careful selection of sounds; those sounds that are most essential towards painting a picture. In film parlance, some of these are "hard effects" -- things seen on screen that have some sort of directly applicable sound effect, like motor noise on a passing car or a gun going off. Some are Foley; these are all the sounds of the characters of the film in motion; the footsteps, the clothing rustles, the fumbling hands, rattle of change in a pocket, etc.

In the film world, these sound are produced by talented, rhythmic and athletic people known as Foley Artists (or, sometimes, Foley Dancers). They perform, like the actor in ADR, in front of a screen, but what they perform is footsteps and small parts and hand tools and bedsheets being pulled and all those other small, usually-human sounds.

So it is a misnomer to say you add Foley to a radio play. You can add similar effects, but the process is much different. Instead of matching to visual, you are trying to substitute for a visual. And there lies the problem. Foley sounds by their nature are fluid and indistinct. They mean something because we see the object that we expect to be making sound. Without seeing a man pull on a sweater, the soft slipping sounds you hear could be anything.

I've found that in general the more concrete sounds work best. Footsteps are great. And then of course what would be "hard" effects; doors, cars, gunshots, etc. You can do some fumbling and some cloth stuff, but it is more like an overall sweetener. Used nakedly, the subtler sounds tend to come across more as noise that snuck into the recording, than as sounds you designed in!

I had a cue for a previous show that was a scuffle taking place just off stage. The artists, taking their cue from the director, recorded the vocals while standing around a table. Dead, dead, dead! I was able to sell some of the scuffle with added sound effects I recorded on the spot, however -- including slapping myself so hard I got a headache!

There's the period problem again; the 50's was light on Foley (modern films are swimming in effects, and the effects are strongly present and heavily sweetened). In contrast a 50's film can be very dry. Even the effects tend to stand out isolated.

Anyhow...I cut a bunch of individual footsteps out of a recording of footsteps on leaves, did some pitch shifting and so forth, and arranged them to suggest some of the blocking that didn't actually take place. But it didn't quite fill the space properly. The effort didn't sound like a film yet. It sounded more like a noisy recording.


I am always leery about introducing music within a musical. In another cue for the same production, I conferred with the Music Director to find out what key the following song began in, and made sure my sound was within that key. This is even more critical when your sound has a defined pitch center and will be overlapping some of the music.

For a full-length movie or more typical theatrical underscore the first composing step is to basically sit at a piano and noodle; to come up with some kinds of themes and motifs. For an except this short, I knew I'd be basically comping; even if a motif showed up, it would be created just for that moment anyhow.

So I put the completed dialog track on loop, plugged in a VST instrument, and started noodling along to see what sort of musical development might occur and what the tempo might be.

Musically, the major moments were as follows; first the girl talks about her encounter with the werewolf. The hero briefly comforts her. Then the Doctor speaks up in one of those "for the benefit of the audience" speeches that in B Movies are often the big morality lecture at the end; "Perhaps Man was not meant to explore space." What I heard in my head for this moment was a french horn or somber brass doing a stately slow march with much gravitas; the "grand philosophical themes are being discussed here" effect.

Okay, and then the switch; the girl reveals the werewolf is her brother AND is a stock car racer (!!!) And to finish up this emotional turning point, the hero notices there is a full moon (apparently rising over the local dirt racing track).

And orchestral scoring didn't work. It probably would have worked if I had had time, but it would have required enough MIDI tracks to write by section and fill out a full studio orchestra; at least three violins, 'cello, base, two winds, keyboard, percussion, etc. And I'd have to spend the time to work out harmonic development and voice leading for all these parts. A good week of work to do it right. Plus of course movie music of the 50's had a particular sound informed both by aesthetics, circumstance, and technical limitations. So more work there in altering the sound of the instruments to feel appropriate and to blend into that distinctive sound.

So the alternative was to score on the cheap; to use as so many budget movies of the time had, the venerable Hammond B3 to comp and noodle through most of the score (with, one presumes, more instruments budgeted for the big title track).

And that also gave me an exciting and iconic way to treat the big turning point; an electric guitar.

Jump back a page. One of the requirements for this effect, stated directly in the script, is "werewolf howls." During the VO session, the director mentioned she did a great werewolf, and demonstrated. Which, since I am a canny and experienced recordist, I captured on one of the mics that was open at the time. With some processing and clean-up that became the werewolf effect for the show.

I liked it so much because of an unexpected quality. This was not a dirty, animalistic sound. There was no slaver in it. Nor was it a mournful, poor-me-I've-become-a-monster sound. Instead it was a full-throated "I'm a wolf and this is my night to howl!"

Which changed, in my mind, the entire character of the movie. Up until the emotional turning point it has been a sad, somber (remember those french horns?) depiction of the descent of an innocent young man into some horrible transformation. Then the wolf howls, accompanied by an upbeat electric guitar chord; this a wolf that revels in his transformation and is not about to be steamrollered by fate. He's gonna howl, and he's gonna win that stock car race, too. If he can just figure out how to get his paw around the stick shift!

So the new version of the score was a mere pedal point under the girl's first speech, then a somber minor-key progression of chords under the Doctor's big speech ("The radiation has transformed him into some kind of a monster, half man, half beast.") And then a jangling electric guitar over the howl of the wolf.

I got lucky; the Doctor's speech worked out to six bars at 110 BPM; I was able to establish a tempo track and turn on the metronome while recording the organ part. The characteristic swell pedal effect of the B3 was roughed in with the volume slider on my keyboard and cleaned up manually in the piano roll view.

But then I went back once again, because the script specifically says "eerie music" and besides just opening with the girl's lightly-underscored dialog wasn't selling the moment -- nor was it making a clear transition from the previous scene change.

So I added a theramin at the top. This is sort of a-chronological; we are joining the movie in the middle of a scene. There is not really a theramin at that point of the score (can you say it isn't diegetic there?) Instead this is like an overlapping sound from the previous scene; at some point before we joined the scene there was a theramin, plus a brassy main title track, and who knows what else. But as we join the scene, that extra-scene element is just fading out.

Well, I think it comes across the way I intended it!

The theramin, by the way, is pre-recorded. I didn't have time to try to perform and/or draw a convincing and idiomatic theramin, and I don't own a real one at the moment. So instead I purchased a pre-existing bit off of my usual supplier.

Anyhow. Last step is to route all the VST instruments to a group bus and apply bus effects to it; a bit of reverb and EQ mostly.

And then to do some overall effects using the master effects section; a fairly strong mid-range EQ, mostly, to make the track pop and to give just a little sense of being a period film soundtrack (I didn't want to go too far in this direction -- the aesthetic concept again of hearing the sound as the people of the time would have heard it. But, also, the track was so nice I hated to grunge it up!)

Thursday, January 26, 2012


So I'm going to walk through some of the steps I go through in making a sound effect for theater.

The vast majority of effects, I find, are not a single sample. Take something very simple; for my first production of Seussical the Musical we needed the sound effect of a pull-chain light switch. I had one, but the actual recorded sound was too abrupt. So I went into magnified view, found one of the "clicks" of the chain passing through the switch, and copied it several times. Adjusted position and volume to make a crescendo, and pasted the switch click back in.

Take another example. Almost no rain, and no water, effect will quite sound right by itself. You usually want to layer a couple to give the sound more depth, movement, and variety.

Take a third example. For something as simple as a School Bell for Grease I had a decent bell sound but I wanted something dirtier-sounding, older and more primitive. So I layered it with a fire alarm bell.

And then there are the more interesting effects. But in my mind, almost every effect is telling a story. It may be a complex story with several characters, or it may be a very simple story with no real plot. But it is almost always a sequence of events that is taking place.

Even if the sequence is as simple as the links of chain in a pull-chain light going through the hole until the switch itself clicks.

First step:
Pull sounds from your libraries, sort them into folders, and purchase or record what you need to fill the gaps.

I have most of the BBC sound effects CDs, and a scattering of other ones. Plus sounds I developed for other shows, and a few things traded and bought and so forth. My libraries are not terribly well organized -- a flaw offset by the sad fact that I have a small set of favorites I end up using over and over anyhow.

Second step: Rough-trim and normalize. This is especially necessary for stuff you record yourself; find the good takes and chop them and save out new individual files.

The example here is the Audacity window from a multi-track music recording session I did recently.

Since we're being basic here, use the waveform view to trim but give yourself a margin -- with a longer margin at the tail (because that is where the natural reverberation lives). Normalization, which is under the "Amplify" heading in Audacity's "Effect" menu, performs an overall amplification of the selection until the highest signal level is at some preset point -- maximum gain by default, but I usually set it to 1-2 dB below max gain to prevent clipping if I add a bunch of equalization.

The advantages of normalization is that all your tracks will have a similar default volume and you can make a rough mix by eye as well as by ear, and the waveform of a normalized track fills the window, making it easier to select edit points by eye. The downside is, besides the potential for clipping mentioned earlier, that you bring up the noise floor exactly as much as you bring up the gain. If you find yourself adding 20dB or more to normalize your tracks, you need to do better work setting your record gain.

Third Step: So I said most effects will have multiple layers. For many effects, one of these layers should be a guide track -- I use recordings from rehearsals for these.

Context is everything. As much as possible I like to have the total picture in the editor -- if there is music in the final setting of the cue, if there is dialog, if another cue is playing, put that sound in the window too. Set it up so you can hear how your sound will seat in the final mix. It is a simple matter to mute these reference tracks just before the final render.

One of the simplest and most useful tool in a DAW or sequencer is being able to draw volume tracks. You can also record fader moves -- and edit them graphically later if you need to. Instead of going through the effort of figuring out exactly how much of a sound you need before hand, import a longer sound file and set fade points by drawing a fader move.

For all those moments when you have to get the timing perfect, here is a little trick; set your sequencer to endless loop and set the loop points to either side of the spot where you need to adjust the timing. Then just keep playing as you adjust.

Fourth Step: Here's where a little mixology comes in. Many of your raw sounds will not be quite right, or not fit smoothly into the mix -- especially if you are asking them to do something unusual. So out with the audio editing tools. These are destructive but conservative; the sequencer creates a copy of the original sound file for each new edit.

The most useful tools in the audio file editing domain are time and pitch manipulation (followed by noise reduction). The modern tools allow you to independently shift pitch or tempo; although a shift of more than 125% either way will start to develop distracting artifacts.

The use of pitch change as a tool ranges from gross to subtle. On the large side, you re-pitch to make a small thing sound large; such as to turn a chirping bird into the cry of a giant roc. Or vice-versa. On the subtle, sometimes a shift of a mere half-step in pitch will help a sound seat better with the sounds around it.

Our ears are exquisitely tuned to cues about the size of the vocal tract. If you re-pitch a vocal up, it will give the impression of being from not just a higher-pitched creature but one with a smaller head. You can't turn an alto into a soprano with a simple pitch change -- you can only change her into a child.

The better pitch changes offer independent control of the vocal formants; meaning you can change the perceived age (size of head for larger numbers) or pitch independently.

For time bending, I frequently shorten airplane and automobile passes. Plays and musicals simply move at too fast a tempo to allow a full twenty seconds for a car to arrive; you need to shorten the time from "just heard" to "motor shut down" however you can; chopping off the start and adding in a fade, doing a tempo manipulation on the audio track, etc.

Tempo changes also changes the feel. I recently did a tempo stretch on a bit of dialog to give the delivery a more deliberate, measured feel.

Oh -- the image here is a fake Doppler Effect using pitch change. For those few that don't know already, when a sound-making object is approaching the sound waves are effectively compressed (each new wave arrives sooner than it would if the object wasn't moving). The result is that the perceived pitch of the approaching object goes up (and the perceived pitch of a retreating object goes down).

The trick to remember is that the greatest change in perceived pitch occurs as the object passes. So you want a curve that approaches vertical as it crosses the zero line, and flattens out on the far ends.

Fifth Step: So this is the sequence; raw sounds and clean-up, assembly, adjustment of audio, and finally the shaping tools for each sound and the total mix; equalization, stereo width, distortion, reverb.

Almost all the contouring tools I use were either bundled with the sequencer or are freeware (a few are shareware). Many DSP companies give away promotional versions with more limited functions -- haunt the audio forums for alerts.

My favorite tools at this moment are: the Fish Fillets from Digital Fish Phones -- a very nice little compressor, expander, and de-esser with an aggressively "warm" sound. On a recent "radio voice" processing most of the distortion was provided by turning up the "saturation" knob on Blockfish all the way.

Eric's Free EQ (no link at the moment, sorry). Also warm and aggressive and very no-frills; three big knobs in sonically useful places.

Vinyl from iZotope -- designed to add a little fake record-player action to a pop track, but at higher levels can help you create a warped scratched old 78.

And of course the amazing open-source mda plug-ins from Smartelextronix; two dozen different useful VST effects with no fancy eye candy, just a full set of parameter dials.

Note of course that here in the VST plug-in world, all parameters from these plug-ins can be automated the same way as a lowly volume or pan. The ability to change reverb depth in the middle of a track can be a real help for nailing down just the right sound.

And don't of course forget channel EQ!

Use effects on a per-track basis, as send effects, as effects on group buses, and as effects on the final master buss -- of these, or a combination of all of the above, can work. Often just a little bit of reverb is all you need to seat the effect.

It is getting late here so I'm not going to go on about world-ization (about using reverb and EQ and distortion to suggest the environment of the sound), but I will add one important note:

Listen to the sound on the speakers it will be played back on during the performance.
If you can, bring your sequences to the theater and do the final tweaking of the mix there. If not, perhaps you can take the final rendered tracks and do a little EQ on them while listening to the results live in the space. There are things you can do in playback, and just changing the ratio of speaker assignments or the position of the speakers can do wonders, but being able to mix in the room is almost beyond prince.

Sunday, January 22, 2012

What is QLab and What Can it Do For Me?

QLab is audio playback software for the Macintosh. It is from Figure 53 and is designed for use in theater and similar live-playback situations.

It is not a mixer, a signal processor, or a Digital Audio Workstation. It is also not a Cart Machine (although Figure 53 has a digital one of those, too.) Also, there is no Windows or Linux version (although there are other, similar software tools). It is also not the only professional-level software for its specific application -- Sound Cue System for Windows is in a similar price range, and of course there is SFX ...of which I will say no more less my blog get shut down by angry marketing people.

With all that out of the way, what do you want this for? Well -- QLab is easy to understand, plays well with all hardware, plays any file the Mac can play (leveraging QuickTime to do so), takes up very little system overhead....and, oh yeah, the basic version is free. And not cripple-ware or nagware, either.

An aside. There's something about shareware. The shareware philosophy is to build a product good enough that people will want to support you in return. The philosophy is to find problems so they can be fixed; which is directly opposed to the marketing-driven philosophy of mainstream software that says the way to sell the software is to slap lawsuits on anyone who dares mention it doesn't work very well. Oh, and a small annoying and productivity-related thing; notice that shareware is usually better about adhering to GUI standards? Commercial software tends to use unusual functions for the control keys, hide the menu bars, and won't even permit re-sizing of the application window. Shareware tends to keep the look and feel of the underlying GUI and that means it is faster to learn it and smoother to use it in a multi-application work flow.

End rant.

So that's 500 words, and what does QLab actually do?

Think of it like a CD player on steroids. A CD player that can play a dozen tracks at the same time, or one after another, at different volumes, panned to different speakers. Now imagine this massive CD player is set up so well you can command it from a single button once it has been programmed.

I've done shows on analog tape, on CD players, on iPods, on iTunes or Windows Media Player, and all of these are relentlessly linear. They want you to play one sound, then diddle around setting up the next sound. A professional CD player or old school cart machine will play the sound then automatically move to the start of the next track and wait for your command.

QLab does this, but with multiple instances. You can have one sound playing as a background, then play two other sounds in a row, then finally take out the long one. Oh, and you can loop, too.

But that isn't even half the power. QLab treats fades and pauses as just another kind of event. So with the tap of a single button, you can fade down one cue and start up another. Or a dozen different events in a carefully timed cascade.

Let's start with the basics. Visit Figure 53's website and download the free version. Again, it isn't nagware or spyware; it won't ask for internet access, it won't ask for your admin password, it won't require you to fill out and return a form.

When you open it, you are presented with a window. The basic format of QLab is an event list. Events are called "Cues"; each cue is an instruction to QLab to do something. Not all cues need to be operator cues; you can set up some to automatically follow others, and you can stuff several within a group and that group itself becomes a cue. And, yes, it can be recursively nested!

The even type we are most concerned with is a sound. Drag a sound file (basically, ANY sound file...mp3, aiff, wav...) into the QLab window and it will become a Sound Cue. Or insert a Sound Cue in the workspace, and navigate to the pop-up to select a sound file to load.

In Version 2 (the current version) there is a black teardrop that indicates the next event that will be triggered when you hit "Go." This teardrop changes to a green arrowhead when the cue is actually playing. When you fire up QLab for the first time, it will select as audio output whatever your default system output is.

Let's demonstrate some of the power of this software right now. Go into Application Preferences and select Audio. You are looking at the default preferences. Every audio output device your Mac is currently connected to will show up as a potential patch. Let's assume for the moment you have something like the 8-output firewire interfaces I love so much. Pull the yellow wire and patch Patch 1 into the firewire device on the list. Now go back to the main window, select that Sound Cue, and select Levels from the bottom row of tabs. You should see eight possible outputs indicated as round yellow crosspoints, arranged in as many rows as the sound file you are using has channels.

Yes...I have installed and used an 8-channel WAV file and sent each channel out individually to an output on the firewire. But more often, I am playing a simple stereo or mono cue through a selected output.

This gives you tremendous flexibility in placement. Even more so when you realize that QLab can recognize multiple audio devices at one time.

Say I have a simple play with a few bits of music for preshow and scene changes, and one cue that is supposed to sound like it comes out of an on-stage radio. I stick a powered monitor under the table where the prop radio is and run a cable back to the booth and plug it into a cheap USB audio output -- such as an M-Audio Fastrack. I then plug a mini-stereo jack into the headphone output of my laptop and plug the other end of that into the main speaker system.

In QLab, I set the M-Audio as Patch 2. Now most of the sound files I bring in will play by default out of the headphone jack and thus over the main speakers. The one radio cue, I go to the Settings tab in the bottom part of the QLab main screen and select Patch 2 instead of the default Patch 1. This sound, and this sound alone, will go out the Fastrack instead -- and show up in the speaker I put under the prop radio.

Next trick. Still in the Settings tab, click the radio button for Infinite Loop. Just as it says on the box, the cue will now play over and over, back to back (how good the loop sounds depends on your loop points -- setting those is simple, but more than I want to get into on this introductory post!)

So how do you stop the cue from playing? Hit the big "Stop All" button at the top of the workspace? There is a more elegant way than that!

What you want to do is create a Fade Cue. A Fade Cue is one of the many kinds of Cue that does not itself contain content (aka a sound file). It acts upon any Sound Cue...any arbitrary Sound Cue, in fact (it needn't be the cue nearest it). Simplest way to set this up? Select your previous Sound Cue and drag it on top of the Fade Cue. It will be recognized and the Fade Cue will now remember which Sound Cue it acts upon -- even if you rename, re-number, or even re-load the sound files.

The simplest settings are to go to the Levels tab and click the master volume slider. Make sure it is highlighted, and down at the bottom. Also check the "Stop target when done" box right beside that master fader.

Move up to the top of the cue list by mouse click, arrow keys, or the Reset All button at the very top of the workspace. Hit "Go." The sound begins to play, and the "next" pointer will move to the fade cue. Let the sound file loop a couple of times, then hit "Go" again. The sound file that was looping fades softly out.

What I have described above will get you through 90% of all shows, or 90% of the cues in a complicated show. What we've created here is a non-linear playback of multiple sound events that is controlled through a simple linear list; merely by hitting the "Go" button over and over (or the space bar) you can start and fade out multiple simultaneous sound events routed to multiple speaker assignments.

And because of the drag-and-drop access, and the total file-type atheism (if it is a valid sound file, it plays. No need to worry about converting to just the right sound format, or even about putting them in a specific directory), if you need two quick songs or even a dozen straight-forward sound effects, you can set them up for easy one-button playback faster than you can open iTunes.

Oh, and did I mention there's basically no lag?

But of course it doesn't stop there. You can set the start and end points for any sound file within QLab, and of course set individual volumes. You can chain cues to automatically follow one another, with offset start times as well. And of course fade cues do far, far more than fade out; they can bring up, cross-fade, or even change speaker assignments on the fly. Cues can not only have names, they have a comments field that displays for each upcoming cue (on those rare shows when I really have time to clean up, I put the actor's cue line and other notes up there!)

And that is just the beginning. QLab also generates and receives MIDI events, can playback video with the same flexibility it offers for audio...

Wednesday, January 18, 2012


You read any good basic description of musical sound, and it will take you through frequency and amplitude, the idealized sine wave, the harmonic series, the way the harmonics can be combined into a complex wave-form. About, in short, the qualities of pitch and timbre that allow us to recognize the different varieties of musical instruments.

And, yes, there is more to say here. There are the so-called formants; specific emphasized harmonics within the human voice that describe the size and shape of the vocal cavity. Within a single voice, the mix of these harmonics tells you what vowel you are dealing with, whether the mouth and throat are open or whether they are pinched. The difference between a sigh and a scream. Within a range of voices, the shift of location of these harmonics tells you the size of the vocal tract, and our ears are so exquisitely tuned we can tell the difference between a deep-voiced child and a male counter-tenor.

And most of these introductory discussions will stop there, with just a little hand-wave about aperiodic waveforms, non-harmonic overtones, and the entire science of how a tone changes over the lifetime of a musical note.

When you get into synthesizer programming and sample manipulating, you learn that the envelope of a sound is as important as the timbre in characterizing the instrument. And you also learn how most musical notes start their life in a chaos of barely-filtered noise before the violently agitated string or the turbulent airflow in a flute settles into the actual note at hand.

You learn about the concept of ASDR envelope (Attack Sustain Decay Release) and the judicious use of white noise and subtle changes in the pitch center to approximate the evolving waveform.

And if we go further towards trying to create a convincing picture of a real musical instrument though electronic means, we begin thinking about non-musical noise.

I learned this quite early; a good acoustic guitar patch becomes that more realistic and believable to the listener if you add a little string squeak here and there.

Anyhow. Shift focus to real instruments, and the reinforcement/recording chain. All the way through the chain, the primary intent is to limit noise; to reproduce just the musical note, without picking up street noises, without introducing odd-order harmonics, without allowing tape hiss or ground hum to rise above the detectable threshold.

Yet, paradoxically, once we've established that clean signal chain we discover that to get a truly great recording we need to include the noise.

On the instrument side, even the apparently simple piano has a whole-instrument sound that is more than the individual notes. The open strings vibrate in sympathy with the struck strings. The sound board contributes its own harmonics. The total sound of the piano is not just a sequence of piano notes, but the sound of a massive iron frame in a heavy wooden box in an acoustic space.

Even further in this direction, the "sound" of an acoustic guitar is not just the notes being fretted, but also must include the many intentional nuances of technique; the hammer-on and off, the slide, the bend. And what makes it truly sing is the flesh; fingers sliding along strings, tapping the frets and pressing against the soundboard.

Or the saxophone, or the human voice; what makes the sound is not just the notes, not just the nuances demanded by the player, but the biological noise of lip and spit.

To get a really good acoustic performance you have to capture some of that sense of real people in a real space. Even if you are creating the echo and reverb of a complex-shaped room with different surfaces and angles electronically, and adding in audience noise from an effects tape.

Of course this varies across styles. Symphonic recordings preserve just a tiny amount of page changes and chairs squeaking. 80's pop preserves practically nothing as electronically-generated tones go through electronic processing and finally arrive at the recording untouched by human hands or any kind of real environment. And small-combo jazz, or folk, is all about hearing ALL the sounds the instruments make (not just the ones indicated on the score.)

Then of course we have the aspect of desired noise. Of microphones that intentionally have a non-flat response because those little peaks and dips better suit the material (such as the venerable Shure SM58 with that pronounced but sweet +5dB peak around 5KHz.)

Of amplifier and compressor stages that introduce odd-order harmonics into the sound, from the so-called "warmth" of an over-driven tube, to the harsher but still potentially pleasing noise of a FET stage, to the all-out crunch of over-driven tubes and clipped signal diodes.

Even reverb can be looked at as a kind of noise. But it is hardly the only "noise" tool in the box; everything from phasing to the Leslie Speaker effect to the crunchingly distorted bit-choppers.

I created almost forty minutes of music for a production of Agamemnon and practically every track included various kinds of distortion and noise, from a little tube simulation to all-out full tracks of radio static.

Here, "noise" is a design tool.

Which brings us at last to sound design. I realized ruefully during one production that practically all of my cues were varieties of white noise; wind sounds, ocean surf sounds, and the like.

But this is a general truth. More and more, I realize that the defining characteristic of most sound effects is noise. There's worldization (altering the sound to make it appear to be organic to the imagined environment), there's layers of indistinct sound that keep the semiotic content of the cue from being too starkly bare, there's sweetening to add attack or low-frequency content.

Oh, I suppose I should explain those more. In re layers, I long ago decided the best way to loop a sound is not to loop "a" sound. Instead loop two or more. Real rain travels in cells of under 40 minutes -- in our experience, it comes and goes, it spatters and retreats. So a good sound cue to play as a background for a ten-minute scene should also ebb and flow, changing its character.

Too, rain or running water isn't a single object forty feet wide. Across the width of a stage, you would expect to hear a little more water-on-leaves here, a little more water-in-the-gully there. So it makes sense to have more than one sound, to place them around the stereo picture (or around how ever many speakers you have available for the effect!) and to have them evolve over time.

In re worldization, of which I've spoken before, the most basic trick is placing the speaker correctly. Use natural acoustics when you can. And don't ignore the potential of bouncing sound; a speaker pointed away from the audience towards a hard-covered flat will create a sound that appears to emanate from that wall.

In electronic processing, the most basic trick after reverb is to pull out the high frequencies for increasing distance. A simple bit of pitch bending (or a pitch envelope) will do wonders towards establishing movement via the Doppler effect. And don't ignore the pre-delay function on your reverb patches -- this sets more than anything else the psycho-acoustic impression of the size of the space.

The old technique is still a good one; if you have an effect that is supposed to sound like it is in a bathroom, a closet, a car...find a bathroom, a closet, or a car. Set up a good-quality microphone, and a playback speaker. This method is even more useful for getting a nice "over the radio" effect to sound right. But I've even used a variation of it by recording on to cassette tape then recording that back to hard disk.

Sweetening goes all the way back to the concert hall. Remember how the characteristic sound of an instrument is as much about the attack and ADSR envelope as it is about the timbre? Well, arrangers for the symphony started quite early on doing tricks like doubling pizz strings with woodwinds; the result was a strange hybrid instrument, a woodwind with a sharp, brittle attack.

On a recent production, I had to create the sound of a whale sneezing. I used myself as voice actor once again, with suitable processing. But to get the impact of the sneeze I added in a snare drum hit and the bite of a T-rex from a purchased sound effect. I lined up these events in the zoomed-in waveform view of my venerable Cuebase, and played with volume until the individual nature of the additional sounds vanished and all that was left was the additional impact.

In Hollywood, a simple smack to the jaw from Bruce Willis may have a dozen different sounds in it, from purely synthesized white noise to a meat cleaver going through a raw chunk of beefsteak.

Anyhow, the insight today is that most of the sounds used in a sound design are more noise than sound. They are aperiodic, aharmonic. They rarely have defined pitch centers, and even when they do, that pitch is not necessarily in tune with any common temperment. And the most useful parts of these sounds are those that are the least like musical notes; the crackles, the impacts, the hisses...all the bits of messy noise.

It is easier to synthesize a decent clarinet sound than it is to synthesize a lion's roar. This is why we still reach for massive libraries of sound effects instead of trying to create them from scratch electronically. And even when a sound is manipulated and assembled and used out of context, why we picked that original sound is because it is edgy, unpredictable, organic. It has hair on it. It has lip noise and spit. It has noise.

Tuesday, January 17, 2012


A rambling post here -- just finished a rather crazy tech and performance schedule (three casts in rotation, each with their own director). As I'm getting my breath back, thinking about that show experience, about my next designs, and also using the time to finally read the excellent Yamaha Sound Reinforcement Handbook from cover to cover, I'm thinking about non-linear artifacts and methods.

The first is in reference to that crazy recent show. A young cast (well, three casts) that frequently went off the rails -- forgetting where they were in a scene, jumping over entire songs, etc. And we had a very technical production with music, sound effects, flying scenery, multiple costume changes, lighting effects, fog, etc.

We were also drastically under-rehearsed (not enough time in one week to fully tech three casts) and we didn't even have decent show communications (well, we had coms, but as a solo show mixer and sound effects operator I couldn't afford to be on headset, and neither is the pit usually on headset).

This meant that we had four different departments having to guess, from experience and dramatic instinct, whether to take or jump a cue -- and where the other departments (not to mention the actors on stage) were going to go. Say the actors appeared to be lost and the scene was stalling. The music director might chose to jump forward to the next song by starting to play the intro. And that meant scenery and lighting had to scramble to arrive at the same spot as well.

It made me want even more of a non-linear system than I had.

There were for that show, as is typical these days, two linear playback systems involved in sound. The first is the mixer. To get through running up the correct wireless microphones for the correct scenes I was using programmed snapshots on the console (Yamaha LS9 series). This was dependent on the actors wearing the same microphones from performance to performance (which they didn't always!) and actually managing to make it out for the scene in question (instead of getting stuck in the dressing room). It also put me in a position where it was harder to work out of sequence.

Whenever I did get the wrong mic, I had to first chase down which one it was. Paging through channel by channel on PFL (pre-fade listen) is slow -- it can take you the length of a song just to find the one mic that belongs to someone not actually in that scene.

You can cheat a bit with the monitors. Given good level indicators (which the LS9 has) you can spot a crackling mic from the meters alone. And it is often clear who is talking loudly in the dressing room and who is singing in a scene -- young actors rarely put as much vocal energy into the latter as they do into the former.

But that almost compounds the problem. If you are just running on faders alone, you can make changes on the fly. But when you add memorized snapshots, you have to anticipate what the next snapshot will do to the evolving picture. And often that means a combination of muting channels pre-emptively, going into scene memory to jump over upcoming snapshots, or even physically holding back motorized faders so they don't go to their memorized positions.

I am at this point unsure of the best options. As has been said, sound boards (in theater use) are a decade behind lighting boards in terms of software. There are options -- sneak, park, track among others -- that lighting boards offer that do not have an equivalent in sound boards. But even those would put you even more in the position of having to brainstorm a complex sequence of commands just to get the following events to happen in the way desired.

Effects playback is a similar problem. This is actually a problem that starts in rehearsal; because rehearsal is all about doing scenes out of order, and doing the same scene (or a single moment in the scene) over and over, you need the ability to quickly change where you are in the playback order.

Add to this, of course, that especially with increasingly technically ambitious shows, and decreasing time on stage in technical rehearsal (because time costs money), that same rehearsal time is often as not development time as well. So your rehearsal playback mechanism also has to deal with changing the order, volume, or even nature of the cues in question.

I'm finding I do a lot of work during rehearsal with QuickTime. I cover my desktop in multiple instances of the QuickTime player, each holding a completed cue or portion of a cue I am developing. The problem is that it isn't as fast as it could be, and it certainly isn't as smooth, to hop between different windows starting, stopping, and changing the playback position and playback volume.

For complex cues in development, I have CuBase open and I play back from there. But this means that in a complex rehearsal I may have to flip between several different applications in order to produce the sounds I want.

I find these workable for trying out ideas quickly and seeing how they play in the space, but less good for trying to replicate a performance over and over again to help the actors in their rehearsal.

Qlab, where the final show is to be installed, is mostly set up in a linear way. You can jump over cues or go back, at least. Qlab also gives you another tool that can be extremely useful; the ability to hot-fire cues by assigning them to a keyboard button.

This can be a bit of a test for the memory, however! A better option is to integrate a MIDI keyboard with Qlab. This, at the moment, is my prime answer to non-linear playback of already developed cues.

First, I assign hot-fire cues to MIDI note numbers. But there is an even better trick; assign a Sound cue to the NoteOn event and loop it, and assign a Stop or Fade cue to the NoteOff event. This turns Qlab into a bare-bones sampler; the sound will continue playing as long as you hold the key down. And you can play as many simultaneous sounds as you have fingers.

In production, I label my MIDI keyboard with tiny strips of white board tape attached to each key. Lately I've been doing this with ideograms; say, a little sketch of a horse head to remind me that this is the key that plays the "neigh!" sound effect.

The second non-linear trick using a MIDI keyboard opens up with Qlab is to go into the Preferences pane and select "use remote control." Then you can assign several keys to play the next cue, jump back a cue, or stop all playback. This is MUCH faster than trying to mouse over to the right place on the desktop (AND it works whether Qlab is highlighted or not).

(This is why one of my Golden Grail projects has been a dedicated Qlab playback controller -- a small surface with buttons that spits out the appropriate play-rewind-stop MIDI events to Qlab, and frees the user from having to interact with the keyboard and mouse of a laptop during the run of a show).

For certain kinds of cues nothing beats live performance. For this, I use mostly a shareware sampler called Vsamp; I load in the sound effects and set up the various mapping and velocity curves and so forth. And this gives me a MIDI keyboard playback with full control over the volume (plus a few other tricks, such as pitch bend for a quick-and-dirty Doppler effect).

For the show I just closed, I added another live playback trick to my bag. The inside of Monstro the Whale was represented by strips of fabric being manipulated by the actors to look like he was breathing. I ran a microphone through my handy Lexicon-200 effects processor (detune effect) and performed the breathing live.

Actually, I wasn't sure I could do a decent sneeze every night, and the pact at which the kids moved the fabric was about to make me hyperventilate, so I recorded most of my performance and translated that into Qlab hot-fire cues. But I still performed some of it live, and the microphone was there in case the performers got REALLY off and I had to quickly do something very different from what was programmed.

And, well, I was going to talk about noise in sound and music, and the way the search for the good sound and the good effect is as much about finding the right noise as it is finding the harmonic content...but this post is long enough.