Invisible

Not a new thought.

But I was listening to radio coverage on yesterday's eclipse, and cicadas came up, and I found myself remembering the process of several sound designs.

Crickets and birdsong are part of that suite of sounds that tell the audience where a scene is taking place, indoor or outdoor, city or nature. They can describe a biome and an emotion; a peaceful meadow or a spooky jungle. And, of course, a time of day (excepting when those crickets or birds get confused by a solar eclipse).

But it is never so simple as just grabbing "birds chirping" from a free online sound library. Besides those little matters of audio quality and legal use, you need to particularize. Chickadees making a racket in a Soho pet shop are not the same as morning sparrows on a farmstead in Illinois. Not all library sounds are equal -- or equally appropriate.

Worse, most plays will have multiple locations, and multiple times of day. For something as simple as Bye, Bye Birdie I had three or four different cricket-and-night-insect backgrounds, depending on the time of day and the emotional mood I was going for; the background to the anticipatory twilight of a hot summer night prior to "Got a lot of living" is quite different from the one-AM still of the final tired chorus of "We love you Birdie."

But that's just primary selection. Then comes the next task.

Library recordings vary hugely in their equalization. Some of the birdsong I reach for frequently have huge (even oppressive) low-end rumble in them. So these need to be equalized and focused in to bring out the sounds you need without masking the stage action and dialog you need to preserve.

(A lot of them also have a huge white noise content. This comes with the territory; birdsong is outdoors sound and wind and background roar are always there. This is wide-spectrum noise, though, and pretty much impossible to edit out. You can only work around it).

Then you need enough so it covers the length of the scene. And starts and stops smoothly, without a jarring effect. Often the library samples you start with are too short. A little over a minute is typical for a library effect and ten minutes isn't unusual for a scene. And you don't want to simply loop (especially because the raw library effect will either start with a bang or will fade in and out leaving an obvious silence in addition to the obvious repetition.)

This is why I like layering several different tracks. I can duck them in and out and change how they line up to present an ever-changing picture that doesn't have so many obvious repetitions in it. (This also plays well with how I like to approach background sounds; to start them louder and more complex and drop them in volume and complexity as we move into the scene proper).

And we're still not done. Because birdsong, particularly, is pretty much defined as being stretches of silence separated by distinct calls. If you leave most library tracks alone, they will be too quiet to make a background track -- except at 1:03, when the bird nearest the microphone suddenly lets out a loud chirp.

I edit the sounds manually. I go into waveform view and "ride the fader" (as in adjust the amplitude moment-by-moment to smooth out the levels.) Audacity has a very nice "envelop" mode for doing this quickly and smoothly.

And here's the trick, and what ties back to the title of the post:

If I do all this, and do it well, no-one will ever realize I did it.



I've mentioned this before. It struck me, while I was listening to that radio program, that the majority of my creative work over the years has been exactly this sort of invisible work.

Hemingway famously tried to write so the words went away and all you perceived was the story. A fellow lighting designer shared with me the conundrum of the box set and the fireplace. The audience will notice a fireplace effect. But if you did your job right, a box set interior (a room in a castle, a hotel suite, whatever) will just look like a room. It will never cross their minds that giving the impression of light coming from chandelier and sconces and or the big bay window is really, really difficult to integrate with having everyone inside properly lit, easy to see, without too many distracting shadows.

(As my friend put it, you spend three days focusing that damned box set to try to control the shadows and the scoop shapes on the walls and so on. And then you plop a $15 fire log from Wallgreen's in the fireplace and plug it in and the audience goes "ooh" and "ahh.")

What little stabs I've taken in the direction of engineering is the same; a good engineering design does what it needs to do in the most direct way possible. A good design ends up looking simple and obvious -- a simplicity and obviousness that was not there when the design process was begun.

With a stretch, you could even include violin practice, since 90% of the job over your first couple of years is the not-making-obvious-mistakes part of it. Your skill is defined largely by how much someone doesn't notice how much (or, rather, how little) skill you have.



But, you know, it really sucks when it comes time to getting credit. For people to mention the sound design in reviews or in the big thank-you-everyone at the opening night party. For reviews, in particular, the only mention is when it went wrong. Never when it went right, because "right" for the bulk of the work is being invisible; the illusion that the singer and the orchestra just sort of showed up in perfect balance and clarity to every pair of ears in the house.

(And since your director and music director and producer and so forth have no freaking idea how much work went into achieving that illusion of transparency, and how fragile it is, they will go behind your back and change stuff and then blame you for when it "for no reason at all" suddenly sounds like shit.)

And it shows up when payday comes around, too.




(There's another example for the bucket; I've been listening to a bunch of archaeology podcasts, as well as the Writing Excuses podcast. The latter is properly engineered. If you just listen to the latter, there's nothing to hear; you just hear the people talking. Listen to the former and you get it; distracting background noises and huge level shifting that makes listening to a dialogue an exercise in daring and frustration as you keep racing to adjust the volume on your earbuds between "I don't know what he said" to "would she please stop blowing into the microphone!")