Sunday, January 27, 2013

Row, Row, Row

So I have a rowboat cue in the current show.

I hasten to add; I'm not happy with that cue.  It isn't what it could be.  But it did get me thinking once again about the process of creating a sound.

In a sort of Aristotelian ideal, you'd discover and define what the actual sound of the object in question was, then construct and refine a duplication of that.  In the real world, jump forward a couple centuries to add a few epicycles; determine what the perception and expectation of the sound is, and discover what is technically feasible.

And then shape for what fits the actual needs of the play.  Which is almost never technical accuracy; instead it is the ability to tell the story, and tell it with the desired emotional content.  And do so within, of course, the total sonic (and visual!) context of the show.

The show in this case is Peter Pan, and the conceit of this production is that the Darling kids (and maybe the kids of the neighborhood?) are acting out all the various situations and character, from pirates to indians to fairies.  The visual for the rowboat, then, is a couple of kids pushing themselves along on a cot-like bed that rolls across the floor.

My conceit for the sound design is this is the world the kids are imagining; so bigger-than-life pirate ships and woods full of strange animals, but with a nod towards the understanding of a child; a non-technical, not-quite-plausble pirate ship.  Or a bow that sounds like a toy bow with those suction-cup arrows.  Only, you know, really cool.

Anyhow.  Start with a recording of the actual thing.  First off, it is nice to find out what something "really" sounds like.  Second, it is (mostly) an easy start-point.  You can find a recording, or go out and make one yourself.    The latter is better because, of course, there is no such thing as "the sound of..." -- sound is complex and contextual.  What it sounds like from ten inches away is not what it sounds like twenty feet away, on a windy day.

But this real recording is also the first place where the ideal falters.  Now, I've actually rowed.  Owned a boat, even.  So I should have memory of what rowing actually sounds like.  But, alas, like so many of us, my expectation is shaped by the sounds chosen by previous generations of sound designers.  I subconsciously expect, in short, for a boat to sound like it does in the movies.

And this is not actually a bad choice.  I mean, besides the fact that those designers knew what they were doing.  Your audience is going to have the same subconscious expectations, and even if you have freshly educated yourself to the real sound, they haven't.  Play them the "real" sound, and they are too likely to go, "What is that strange sound we are hearing?"

The other problem with real sounds is that they are dirty.  They are complex.  Often they have other elements in the recording, often there is recording noise, often it is too short or pans wrong; but even beyond that, the real sound of a real thing contains too much information.  Too much is happening.

In the movies, real sound off the sound stage is rarely used.  Instead, the sound field is reconstructed, footstep by footstep, jingling coins in the pockets and the slither of a silk scarf and creak of a leather jacket and all.  What is created by the dubbing mixer, with the aid of the Foley stage, is an illusion of natural sound.  It is sound with the extraneous cut out, sound that is focused on those elements most needed to tell the story, sound that is shaped artificially to elicit emotional reactions.

And we often need to do that in our stage sound effects as well.  Instead of the complexities (and lack of instant recognition) of an actual recording, you shape, you rebuild, you sweeten, and you fake; to create a model of a sound that sounds like what we expect it to sound like, and contains only those elements it needs to.

So model an imaginary rowboat.  My mind expects two things; a squeak and a splash.  There's the squeak of the oarlock.  Then the splash and swirl and drip of the oar entering water, pushing, exiting dripping.  And a little bow slap and some minor wood creaking.

Actually, most oarlocks don't creak.  What they do more is rattle.  The recording I used for this show was mostly a dull wooden rattling, and no water noise at all.  I spiced it up by adding some water noise.

Because of the dual aspect of this cue -- that it both was a real rowboat and was kids playing rowboat in a backyard -- the water I added was a BBC track of someone taking a bath in a tub.  So smaller splashes, with that close bathroom echo.  And I EQ'd the boat mercilessly to haul out as much of a squeak as I could and downplay pretty much all of the other sound.

Had I time, I would have constructed the boat bit by bit, with samples that had nothing to do with a rowboat.

Which is how I made the pirate ship.

Again, I've been on real ships.  I have internalized the kind of creaking the wood of a large wooden ship makes (among others, I've been on The Golden Hinde.)  Also spent enough time on small sailing ships and around marinas to internalize the flapping of sails, the lashing noise of rigging against masts.

My "visual" conceit is that the ship is revealed as if in full sail, at that moment when the sails fill with wind and snap taut.  Of course the entire point of Captain Hook's ship is that it doesn't go anywhere, and probably has the sails furled as it lies at anchor in a sheltered bay.

So the primary elements were the body of the ship creaking, some creaking ropes of rigging, a little water lapping against the hull, sail cloth flapping, and of course the big "whooomph!" of the sails opening right at the start of the cue.

The creaking was a library sound effect of a wooden rocking chair, down-sampled and down-pitched.  The sails were from a BBC recording of the sails flapping on a small sailing dingy as it is coming about into the wind.  I've used it before, but this time used it almost at pitch to make the pirate ship slightly more toy-like and accessible.

The rope I re-used a sound I recorded a long time ago; I tied a knot in a piece of old hemp rope, tied it around a wooden ladder and recorded it on to my mini-disk.  I've long since lost the original recording, and what I have made a long side-trip through the 12-bit memory of a W30 sampling workstation, but it is still with me.

The lapping was water against the shore of a lake.  For this show, I made a few new sail sounds by setting up an omni microphone on a quiet stage and running around in front of it with a bed sheet.  That amused the stage manager no end.  Unfortunately the stage was not quiet enough, and by the time SoundSoap had taken the noise of the stage light dimmers out of the recording it was almost unusable.

But only almost.  One of those snaps of fabric still begins the completed sound cue.  Oh, and I added some gulls and surf at the start of the cue to "sell" it.

This is a very, very useful technique.  You want the sound of a train?  The audience might get confused by the puffs of a steam engine, and truly lost by the pound of a modern diesel, but add a train whistle/horn...

There are a great many cues that you can add one extra sound that will explain the cue to the audience.  Because, after all, they aren't visualizing what you are visualizing, and their auditory memories and associations aren't yours, either.

The technical factor that has hit me every time I have done a similar ship sound (this makes about four times I've build a similar bed, though each had a unique character), is that you need to have something going on all the time in order for there to be, well, a sound playing.  But many of the elements are periodic and/or distracting.

So the final balance between all these different elements; lapping water, creaking wood, flapping sails, etc., is as much determined by the need to make a relatively smooth overall level; a sound that tells the story almost continuously, without rising in volume so much as to mask dialog or even as to attract too much attention, and without lowering enough in volume at any moment to attract attention by its absence.

And here, again, constructing a sound from little pieces like this makes it easier to make a sound that has this quality of chunky peanut butter.  Of not being too spiky, or too bland, but being of a consistent texture that can be extended for as long as the scene lasts.

No comments:

Post a Comment