Friday, August 30, 2013

Not-Working at the Computer

It has become the pervasive image of "working," now -- someone sitting at a computer.

It is so straight-forward and seemingly sensible.  You make some coffee, sit down at the desk, boot up the computer...and do work stuff.

And, yes,  I have a lot of work that is done on the computer.  Mic breakdowns.  Programming.  Writing sound effects.  Organizing calendars and communicating with email.  So this isn't an entirely wrong assumption.

I have right now a bunch of things that are coming up on deadline.  Some are shows or rentals, some more personal projects.  And as a result I have a quick breakfast, make a big mug of coffee, sit down, boot up.....and accomplish absolutely nothing.

Because even though a computer is part of what I need to do, the computer is not all of it.  I program on embedded hardware and without the hardware I can make little progress.  I'm working out mic plots and sound plots and effects for upcoming shows and these have to relate to the script, and to rehearsals -- neither of which is on the computer per se.

Sure, I could bring out the script, the hardware, the other tools I need.  The problem is one of focus...the instinct was to sit down and WORK, and going through drawers sorting hardware, or marking up script pages, isn't as straightforward as, well, booting the computer.

And the trouble is, by the time I've done the emails I have to, and checked the calendar, it seems sensible to update the blog and see what is up in forums.  And then I'm a little bored and the coffee is half-gone and I take a break by looking at some web comics.  And before you know it, half the day is gone and I've made no progress on the pressing work.

Of course by the time I've made that realization, I've been sitting at the computer all morning, and I need a break.  And some lunch.

And thus the day passes.

Thursday, August 29, 2013

Sound 101: The Mic Check, For Real

I couldn't quite organize the last entry to describe real, practical mic checks for wireless mics.  Hence this one.

IF you get the chance for a proper mic check, this is what you want to have;

The actors should be wearing the mic.  Some will try to come out to mic check holding up the microphone with their hand.  Send them away until they've taped it on.  If the actor has a hat or mask or wig, that should be on the actor as well.  If the actor is wearing metallic fabrics, those should be on their bodies.

The first thing to listen to is silence.  When the actor isn't speaking, is there hiss?  Then you may have the sensitivity too low and the amplification too high.  Or a bad transmitter.  Do you hear an electrical whine or warbling?  Bad transmitter.

If you hear a "whiff" or the sound cuts in and out, put an eye on the RF strength indicators on your receiver and have the actor walk slooooowly around the stage.  If you have the time, or you suspect an RF problem anyhow, it is worth having the actor visit every part of the stage they will be using and make sure you get adequate signal strength every where.

If you hear a rustling, crackling, or popping, suspect a bad connection.  It might be anywhere in the system, from the pot on the head amp to the jacks in the patch bay, but we're worrying about mic test now.  It's probably going to be the element.  Have the actor move around.  Send an assistant out to the stage and have them tug on the connector, twist the cord, and otherwise jiggle things to see if you can make the sound worse.  And send your other assistant for a spare element!

(For stuff like this, it is far better for everyone if you just replace what is on the actor, mark what was on them as suspect, and do your diagnostics later when you don't have a room full of people standing around waiting on you.)

Second thing is voice.  Many actors have been trained by many (poor) sound checks and will go straight into belting out a song -- usually off key.  I try to have them speak instead.  I learn more about the natural timbre of their voice when they speak.  In a stage voice, mind you.  This is one place where "indoors voice" is never appropriate.

Assuming you previously dialed in the desired sound and placement, you are now listening for two things; that the voice still sounds like it should, and that it is free of new artifacts.

Hollow, flanging sounds can come from hats, from the placement, and from something as simple as the tape slipping over the head of the element.

A heavier, more phlegmatic sound than you expected is almost always an error in placement.  Left to their own devices, actors will tape lower and lower on the cheek, and closer and closer to their mouth, as if subconsciously trying to save themselves some of the effort of singing full voice. 

An oddly distant and often weirdly tinny sound is often the result of the mic slipping entirely, ending up inside a wig cap or behind an ear.

A thick, wooly, or chocked sound happens when the filter cap needs make-up cleaned off it, and as elements corrode and age.  You can often bring a little clarity back to Countryman elements by washing them out in rubbing alcohol -- and making sure they are completely dry before you power them up again.  But they will inevitably lose their tone over time, sounding cheap and muffled after a year or two of use.

Most actors have a different sound in their singing voice.  It is well worth having them open up.  They should be singing at least something similar to a song from the show, and in a register they use often in the show (again, for some peculiar reason actors will chose to do sound check with the one verse they sing falsetto, or in some ultra-peculiar character voice).

Check -- assuming you didn't check before -- that the gain staging is clean and they aren't clipping.  And listen as before to make sure the sound like you want them to sound.  Be very aware that actors will make mic check before they've warmed up.  Also, a significant number of actors can't find their pitch without aid.  When they sing a Capella, they will sing in the wrong key (or between keys) and they will find it uncomfortable and strange-feeling, and as a result they won't sound the way they do in the show.

Which brings me to the biggest caution of them all.

Over and over, I've seen this done in mic check.  The mixer asks each actor to step to center stage and start singing.  Then he adjusts their EQ to sound "right," and adjusts the head amp until everyone doing mic check that day is roughly at the same level.


Yes, you want to EQ.  Yes, you want to adjust your head amp so you can put the faders for the ensemble straight across.  That way, you always know without consideration of which actor it is where your average dialog and song settings are, and where the faders need to be to achieve a rough vocal blend between several actors at once.

And you should do this the very first mic check you have, because that is the only time you will really have the luxury to listen to just that voice alone.

But just as in mixing, you don't want the perfect sound for that individual instrument.  You want the sound that works in the mix.

And, as I said, the actor in mic check isn't doing what the actor in the scene will do.  So when you get to their big solo number in full dress rehearsal, well, then you do the final adjustment to zero in their sound.  Their EQ, their compressor settings, and of course their gain.

Don't second-guess yourself during mic check.  You had the settings right in last night's performance, so don't mess with them!  Adjust only that which has changed, or was clearly a mistake.  Otherwise, understand that the actor is cold, bored, out of key, and not wearing all the parts of his costume, and two hours later backed by the full back and across the stage from the soprano he is going to sound different.

Be conservative.  Use mic check to confirm that everything is working normally.  Don't sit there and try to re-write the show.

So what if you can't get a mic check?

First, check the gear before the show.  If you have problematic gear, bring each and every mic down to the stage and test them yourself before you hand them out.  (This is why us sound engineers don't have day jobs.  Our pre-show checks need to start too early in the evening).

Especially if this is the first test of the gear, be thorough.  Listen to the silence.  Talk into it.  Wiggle the cord.  Shake the transmitter (best way to discover bad battery clips).  Walk around the stage looking for holes in the RF.

At my current space, I do all the check I feel is necessary while I am prepping the mics.  When I've got a loose wire, I can see the "clip" light blinking as I stick the condom on the transmitter.  I also look for problems in the display, the battery charge, loose connector, or anything else that stands out.

When I get to the board, I check the tell-tales on the receivers.  They should have strong and characteristic RF, and show some activity in AF from the general noise of the room.  One error that leaps out is if the RF is low but there is a ton of audio.  That means the mic isn't working at all, and I'm seeing a television station in the same band.

I love the channel strip meters.  I should see the bottom indicator lit.  If I clap my hands near the shoe tree where the mics are stored prior to the actors getting them, all of the channel strips should light in about the same amount. 

A signal that violently goes to very strong and back to normal, very quickly. is either someone dropping the mic on a counter, or a bad connection.   So is a signal that stays at a high level but steady, without moving.  And if the signal goes dead, once again we have a dead element or a dead mic.

When the mics are finally on actors, then, I can listen via PFL and make sure the actors sound normal.  Unfortunately they are backstage, you are on headphones out in the audience, and there's no way to make them talk or sing on cue.  Vocal warm-ups are good for this; if most of the cast is warming up, you can zip through most of your mics and confirm they are being worn correctly and are still working right.

The meter check and the PFL check also take place during the show.  Particularly in kids shows, or early in the dress rehearsal process, where you can't be sure who is actually singing and who is still back in the dressing room.  I've gotten pretty good at recognizing the difference in sound, and even in how the channel meter strip looks!

Monday, August 26, 2013

Sound 101 : The Mic Check

What is that Huxley quote about a beautiful theory being slain by ugly facts?  Mic check for wireless mics is a perfect case of something that seems like it should work and is sensible, and in reality tends to work out poorly.

Every time a mic fails in front of an audience, someone -- director, stage manager, producer -- will be asking "Can't we have a microphone check before the show and stop this from happening?"

And that's the lovely theory.

Even more attractive, from the point of view of the sound professional, to be able to do a full sound check with the actor...!

So here's what this lovely image looks like;  each actor is fully dressed in the microphone and the costume they will be wearing, including hats and wigs.  The room is silent, the actor warmed up, and there is time to probe every nuance of their voice and that fragile wireless link to see that the equipment is working perfectly and that the adjustments -- compromises -- we are making from position on the actor to EQ at the board are the best choices possible for that role and that space.

Every bad connector, patchy wireless path, metallic costume, poor position, corroded element is discovered and the problem repaired.  The slightest noise or stale sound means fresh equipment is brought out...and an hour later, the shows unrolls with flawless performance by the microphones and not a single failure.

And here is the first ugly fact in that beautiful image.  Mics don't fail in the box.  They don't fail in the bag before the show.  They fail on bodies.  They fail because they are being sweated upon, wrenched at, jiggled, sat on, and otherwise tortured by being taped to an actor who is dancing and singing and sweating and having water thrown on their face and buckets put on their head and all the other things that befall them during the course of a show.

A significant number of mic failures -- particularly the loud, uncorrectable kind -- happen suddenly without warning.  They happen when a cord caught in a costume, or a drop of sweat rolled into an element.

And no check prior to showtime will catch them.  Heck -- they will make it through four scenes and three songs flawlessly.  And then die....usually right at that very moment the actor exits the wings and will not leave the stage again until the final curtain. 

And they fail in the context of a show.   They fail because the RF can't punch through the chorus, or hetrodyne interference crops up when one specific trio happens to be in one specific arrangement on stage, or at 8:15 PM when the local pirate radio station fires up, or in the lighting cue of Scene 5 when hum in the wiring to a set-mounted light is picked up by a mic.

Still, there are all those problems of corrosion and fraying that will show up slowly, over time.  And most brands of element decay in sonic quality over weeks of being worn and you can hear that sonic degradation slowly increasing.  And transmitters weaken and start drifting off frequency.  And all of this, you could catch in a mic check.

But here's reality in all its callous glory.  That mic check?  Half the actors will not be in full costume.  Some of them are outright lying to you; they have tacked on their microphone just to get through the nuisance of mic check and will remove it promptly the moment they are back in the dressing room.  (The bolder ones will hold it up with one finger as if holding a mic in the vicinity of their face is equivalent to having it properly taped on).

The band is tuning up and even with headphones you can barely hear.  Not to mention they are vacuuming the lobby behind you.  You don't have time to listen to each voice; the set people need the stage to set up, the front of house staff is clamoring to open the doors, and there's a dance brush-up and fight call that are also competing for the time.

The actors are cranky, their voices cold.  No-one is there to give them pitches and they will sing off-key (even when they sing a number that's actually in the show). Which means their vocalisms won't much resemble what they will actually be doing on stage later.

All of these can be addressed, if you have the time and patience and clout to fight it out.  Which is also time and capital you are going to want somewhere else; every time you go to the mat for something, you have to give up something else in return.  You have to pick your battles, and in the scale of all things sound, I choose not to fight the actors and make them my enemy over some of the details of mic check.

And, yes, in most theaters you don't have the luxury of plentiful spares.  Even in relative affluent circumstances there are sheer limitations in just what you can stock and how many frequencies can be crammed into the narrowing airwaves, and all too likely a potential problem is going to have to be signed off on in favor of the time and crew and money necessary to fix a worse problem on one of your leads.

Depending on budget and time, it can get a lot less like the final polishing of a spacecraft about to launch, and a lot more like triage.

Which is why I often forgo mic check entirely!

 I know.  It sounds ridiculous.  In the best of all possible worlds I would have one.  At my current theater, we have a 7 PM curtain.  Practically speaking, the only way to get the cast into microphones in time for a proper check would be to demand they leave their day jobs early on every performance night.  There simply isn't enough time before the house has to open.

But I check the mics on bodies.  When things are scheduled properly (the last show, they were not), mics are on most of the cast by warm-up, and this is a perfect opportunity to listen to full singing voices.  The only difference is, instead of playing it over speakers I am doing PFL into my headphones.

The meter strip is also a good check.  I have the receivers near me at FOH, and I can check battery status and RF strength of every transmitter from there.  I can also tell from the meter strip if an element is shorted out, or gone completely mute. 

All of this would go much, much better, though, if my cast would get into microphones earlier than the pre-curtain speech!  And that is the main reason I'm contemplating renewing mic check.  Not because the check in and of itself will help significantly.  But because forcing the cast into microphones with plenty of time means I can do the checks that matter, and do the repairs that are necessary, before they need to be on stage.

Thursday, August 15, 2013

Rust and Steel, Cotton and Code

You never quite forget skills, but boy howdy...they can get rusty.  I've been looking over my new Bernina and contemplating projects with it (as it is far too good a machine not to be actively used.)

And I see I've forgotten so, so much.  Either that, or where I got when I was studying costume design and construction was a lot less far than I thought it was.

I was just looking up how the Overlock setting works, and that led me to Overlock Feet, and that led me to the bewildering variety of presser feet available, including the seemingly very useful zipper foot, which also has something to do with Invisible Zippers...

And I don't remember this from when I constructed the Emerald City Chorus back in High School.  I vaguely remember seam tape and maybe a French Seam somewhere, and I know I never did learn the Serger, but the rest of this feels like unknown territory to me.

I remember taking measurements off cast members.  I remember adapting patterns.  I remember working in muslin and doing fitting, I remember stitching in armsceye or pinning breast darts.  But I can't quite remember the details.

And it is obvious, on only a cursory read, that there is so much about fabric types and machine techniques and varieties of seam treatment and so on and so on that I never ever learned.

The only totally for-me thing I have sewn in the past several decades is a pillow slip I just created from a cute Japanese-import cotton print.  And, yeah, my confidence is unshattered; I still think I could make a jacket if I chose.  And my general realism is intact; I know it would be a project of a week or more.  What I'm starting to realize is why it would take that long, and how much of that time would involve learning basic techniques and discovering common pitfalls.  Oh, and also seeing some of the ways sewing can turn into an expensive hobby.

And of course, at nearly the same moment I found an overlock foot on eBay and was sorely tempted, I also discovered I am eligible for a free one-year membership at Tech Shop.

Which would be a chance to finally get some time on milling machine and metal lathe.  As well as access to various metal, wood, plastic, and so forth machines -- restricted mostly by the waiting list, and the need to take the safety and familiarization classes for each machine in turn.

Except, as with the sewing machine, where I've been is in a place where every now and then a project would go better if I had access to so-and-so.  I'm not really in a place where I have a project (or series of projects) that demand frequent access to specific technologies.

Sewing included.  I made a new set of mic bags, and repaired some items around the house, and that basically finishes up the outstanding projects.

My current external task list is to clean up some vacuuform drum magazines, create a 3d-printable model of a Cadillac-Gauge V150, and create a mock-up receiver for a de-militarized Suomi.

On my personal tasks, I want to put more 3d models in my online store (which is now earning about twenty bucks a month but could do much better), and create the (mostly software) infrastructure to make my Duck Node concept a viable theatrical tool.

None of these actually require a lathe.  Or a sewing machine.

Wednesday, August 14, 2013

The Geek Cred Check

Yes, it happens.  Male geeks do it to each other constantly.  It's part of the fun (for at least some of us.)  The problem is when it is used to be exclusionary.

And I think this is a venue thing.  Makers Faire?  Not cool.  It doesn't matter if the person you are talking to has never held a soldering iron or a Dremel, or is Harrison Krix.  Don't do it.  Don't exclude.  Don't assume.  Don't talk down.

Same thing for anyone -- for everyone -- at a convention, be it Dungeons and Dragons, Comic Books, or whatever.  They paid the price at the door, they made time on their weekend to come out?  Then they are part of the club.  That is what that venue is there for.  It is for everyone to enjoy, regardless of whether they can name more than one Green Lantern.

But way off in another corner, there is tech.  It is one thing if you are sizing up someone just because they happened to mention Arduino in a social setting.  But it is another thing entirely if you are working, and the quality of the show may depend on that person actually knowing what it is they claim to know.

Every place I've done audio, or theatrical lighting, I've had to play the lock antlers game.  I've been the one defending myself just as often as I've been the instigator, too (I tend towards a lighter touch when I'm probing a co-worker's abilities.)

I need to say more here.  I've been on a job more than once when a hotshot has tried to take over, shunt me aside with a, "Since you obviously don't have the l33t tech skillz I do, we're going to do it my way."  And that's when I go into full riposte, pulling up tech specs and factinos and, if I have to, cutting them off at the knees in the process.

And, yeah, my ego is what causes this, but I also have a belief that I know my stuff and know what works in my theater, and it would be bad for the show for someone who is only trying to show off to go around changing what works, fixing what wasn't broke, and basically putting us in a place where we can't get the show up on time.

Especially when the antler game reveals a deep and entrenched Dunning-Kruger on the part of the person who is convinced the best thing to do is rip it all out (without even looking at what is already there) and start from scratch the "right" way.

And, yeah, this sort of locking antlers can be fun sometimes.  I've done it over tech, over science, over the few SF or comic book properties I'm actually interested in (not that many, really).  And even over history -- my best friend is a history buff with an incredible memory and I just HAVE to throw in a, "The world wonders?" every now and then just to show I'm active in the conversation.

As I'm thinking about it, I think part of the deal is that geek cred in a specific subject is not the same as having basic geek creds.  It's like a one-point skill versus the 10 point all-skills. And the problem is, if you are a white straight guy who isn't too obviously into sports, you pretty much get the basic cred for free.

I can cheerfully admit I stopped reading Marvel just about when Power Pack moved to NY, and to me Jean Gray only came back once (as the Phoenix).  I can ruefully admit I don't grasp object-oriented, and I only do a little bog-standard C.  I can own up to never having hooked up a ribbon microphone, and having no time in a real studio.  But none of those admissions make me not a credentialed geek.  They just mean I don't get credentials in that specific genre.

And this wouldn't be the same if I were female, a person of color, or some other convenient "other."  Where the assumption of geekitude isn't default, and I'd have to do the antler thing just to be accepted as having an expertise in something.

And then, of course, since the entire point of the exercise is exclusionary, sorting the us from the them, were I that "She's one of them fake geek girls!" person failing even one of the spot checks is counted as failing the entire exercise.  As that pre-supposed other, I would not be allowed to shrug off a question with, "Never read DC, never got into MLP, but I know the Whoniverse up down and sideways."

And, you know, I'd like to think the tech and Maker communities are better than this.  That we could instead embrace the diversity.  "You don't solder, but you know how to sew?  Hey...maybe I could learn something from you."

But it has to start with not doing the credential thing anywhere other than on the shop floor, where power tools are spinning (or on the grid when a show is opening in three hours).

Or, rather, not doing it in a, "Prove yourself to my standard or I'm going to kick you out of the club."

Monday, August 12, 2013

Almost Perfect

The WIZ jacket finally failed in performance.

I'd been making a repair about once a weekend.  Each of the ballast resistors came loose once, and several of the solder joints to the flexible LED strip on the collars broke, but until Sunday matinee it had a flawless run.

I am still deeply satisfied, because I am not a little surprised.  It was thrown together so quickly, and I never had the in-depth testing cycle I wanted.  I didn't honestly expect it to make it through Opening Weekend without flaws.  And it failed in a good way; it didn't turn on.  If it had failed to turn off, or if the lights had gone dead along one side, the audience would have realized something was broken.  As it was, the scene was merely a little less magical than we had intended.

I'm not 100% sure of what went wrong, and this is because my procedure was diagnostic, not forensic.  When I looked at the coat, the 9V style battery jack had pulled entirely off the battery pack.  Presumably the dangling length of battery cable had gotten snagged when he was dressing for the scene.  I've taped that now, and tacked up the dangling cable as well.

But when I plugged it back in, the coat still didn't answer the controller.  The picture above was taken by a co-worker after I'd pulled all the equipment boxes out of their pockets in the back of the coat and was probing at them. 

Blinkenlights on the Arduino and the XBee Shield showed they were getting power -- so battery cable, that part of the wiring harness, and the power regulators all checked out.  When I opened up the power switch box (the Altoids tin) I confirmed power at the busses there as well.

I plugged in a USB cable and via the Arduino IDE looked to see if I was getting a serial response out (there are several serial print commands written into the coat's software for debugging purposes.)  Nothing came out.  I reloaded the Arduino software.  Still nothing.

Rebooted the control software on the laptop (a Processing application that talks to the XBee transmitter on a Sparkfun USB breakout).  At that point, the Assoc Light lit up on the coat's XBee module, the coat confirmed reception of controller commands by echoing them back through the Arduino IDE...and the lights lit.

So my best guess is that simultaneously, the Processing ap borked (possibly because someone had jiggled the USB connection and it didn't restore it properly), and the plug pulled off the battery pack in the coat.  But I can't rule out that the software on the Arduino went bad (this does happen) and it needed to be reloaded before it would work again.

Wasn't a waste of my lunch break, though.  I took the opportunity to tack down the bags holding the equipment boxes, and tack down more of the wiring.  I've been tempted to take the whole thing home and re-stitch the equipment bags and add proper velcro closures...but it works, and it made it this far, and we've only got two weekends to go.

Friday, August 9, 2013


With the arrival of the sewing machine, my tool collection has grown once again.

Problem is, I have no space to deploy what I have.  I'm trying to put a small workbench together to hold the Dremel Scrollsaw, the Central Machinery (bah!) Drill Press, and the woefully underpowered Ace Bench Grinder.  At the moment, I'm using the Bernina 830 (the 57 Cadillac of sewing machines) on my drafting table, which also holds computer, the Krytronic Soldering Station, and serves as dinner table as well.  Oh, yes, and sometimes I draft there as well.

I dream of a lathe.  Doesn't everyone?  Not that a CNC mill would be a bad thing either.  Or a laser cutter/engraver.  I'm happy enough with sending out for 3d prints, though!  Some day when I'm feeling flush I mean to pick up a membership at the local Tech Shop, where I can finally get some lathe, mill, and shop-bot time in.  Welding, too, is something I'm unlikely to do much more of at home.  I reserve the right to braze some more brass and copper, though.  After all, I have the gas, and a box of fire bricks...

So here's a couple of other tools that have caught my eye of late;

Cordless soldering iron.  The last show was informative about how much of a fuss it can be sometimes to get power to a soldering iron out to where the work is.  I'm not sure this is quite recurring enough to carry a cordless iron with me everywhere, though.

Router attachment for the Dremel (the rotary tool, not the scrollsaw).  I have been itching to make a replica prop with the MDF technique, and slots and panel lines of those really call for a mini-router.  The router attachment Dremel makes is horrible.  Stew-Mac has a much better one but the base kit is $54.  Another tool I may have to wait a while on.

Wood/mini lathe.  When I say it is smarter to save metal working for Tech Shop, that doesn't mean there isn't a lot of useful stuff I could do with a lightweight lathe with a mere four inches between centers.  There are however very few in this category, and most of them are seriously underbuilt; you are better off constructing your own.  Which ends up being a few hundred bucks in parts.  Also, the thought of dust control, in my small space, is a nightmare.

I keep meaning to have a duplicate set of electronics repair tools so I don't have to pack and unpack my gig bag.  And now that I have a machine of my own, I could make a nice little tool roll for them as well.  Of course me being me, the temptation is far too high to spend the money on some silly fabric for it like a Sailor Moon print...

Especially because I have two different resources now for custom-printed fabric.

Ah, that's the problem with tools.  When you have a hammer, you start finding nails everywhere just longing to be hit.  Right now, everything looks like small flexible squares of woven material.  I've already loaded up a second bobbin, and I'm reading up on embroidery stitches...

Wednesday, August 7, 2013

LiPo and Buck-Boost and Xbee and Cree, These Are a Few of My Favorite Things...

I'm in production, working a 30-hour week for show calls alone (plus ongoing repairs and maintenance, meeting on the next two or three shows, etc.)

Plus I've really fallen behind on such things as house cleaning and car repair.  Plus finances are still not good, meaning I need to concentrate on work that pays instead of long-term development.


An upcoming show is going to use a lantern.  They seem happy to handle it themselves -- didn't even think of asking me -- but is a good excuse to continue the development I started back with my BlinkM experiments.  Especially since The Wiz showed I could make a battery-powered remote-controlled rig that PWM'd multiple channels of high-power LEDs.

The BlinkM board from ThingM is very cute; it combines a "Piranha" LED with an ATtiny controller.  They package it with an easy-to-use software that uses an Arduino as a programmer (or you can use another USB to serial converter, like my own favorite USBtiny ISP from Adafruit).

But it is underpowered for stage use.  Especially if you intend to actually light part of a scene with an oil lantern hanging in a barn, or a storm lantern in an actor's hands -- or something more unique like a glowing crystal ball or Wizard's Staff or whatnot, we really need to reach for the power of the Cree; several options in 1W LEDs, single-color and RGB.

Unlike the mere 30ma a channel of a Piranha, though, the 350ma of a Cree can't be switched by a naked AVR.  Plus the power-hungry nature of these things means you really don't want to be throwing half your battery life into cooking a 2W current limiting resistor.  The best option is an actual current-regulated supply, better yet, a buck converter that trims the voltage down to a useful level without throwing half the power away as heat.

The reason for all this complexity, of course, is that with a bulb and a wire (or even a high intensity LED and a wire) you are limited to static display.  With the embedded AVR, you can flicker like a candle, or pulse or change colors.  You can dim up and down instead of the light switching instantly to full power, which also makes for a more realistic effect.  You can adjust the color of the effect, and the intensity of the effect, without having to wrap gel around it.

And this level of control can be achieved remotely.  In the case of the WIZ's coat, via a laptop in the booth.  The simplest link (and one of few with the range for theatrical use) is the XBee modules.  Unfortunately they are also not that cheap, and have other electronic requirements.

Which brings me to the tangential "Duck Node" concept; small battery-powered XBee-enabled devices that can either be used to sense an actor's motions or other commands, or reversed to control an effect -- such as the battery-powered high-intensity lights we are talking about.

What I want to build right now, then, is a prototype lantern.  For the upcoming show, it would work just as well with a mono-color LED, a switch on the side so the actor controls it, and not even a flicker circuit.

But I'm going to give it the full bells and whistles; RGB LED, and full remote control.  The only trick at this point (since I have all of that working now on stage in another form factor) is getting it to fit inside an oil lantern.

If the prototype works out, though, I'll be able to custom-design the next circuit to be printed instead at a fab house.  And then I might actually have something to bring to the next Maker Faire....

Slowing down from 90

Worked a 90-hour week last week what with tech and two different shows performing, plus change-over, plus the usual ongoing show maintenance.  Might have been a little less, might have been more; all of it was contract so I didn't keep an actual time card.  But most of the days were 9 AM to 12 PM with maybe half hour to eat once in the middle, so....

Was expecting to collapse when it was over but it hasn't happened yet.  Although today was a little less things accomplished.  Yesterday I deposited checks, dropped off mail, did laundry, read half of the Reaper manual, went to the gym, cleaned the bathroom, replaced the shower curtain, repaired a futon cover, shopped for tools and updated my blog.  Today I had one meeting (well, it was a two-hour one), cleaned the refrigerator, bought groceries, pulled the radiator fan from my car tested it and ordered a new motor online...but otherwise rested.

I still feel pretty good.  If I could just hold on to the kind of productivity I had last week, I might actually get something done every now and then.

Tuesday, August 6, 2013

Non-Linear Playback; The Evolving Story

Seemingly all the sound effect designs I do these days have some non-linear elements.  But at the same time, the majority of the show is still effect by effect; often called and/or executed by the Stage Manager, and presented in one long list.

"The Wiz" (which is running now) is one of these standards.  My Sound Assistant is on headset, taking each cue from the Stage Manager.  I've built in dips (fade cues and restore cues) to take down the sound effects during songs, and these are also being called as lettered cues by the Stage Manager.

The one exception on this show is also the first time I've had someone other than myself executing an improvised effect.  To wit; the wind effect for the twister is meant to be artificial, to be more like a performance on a synthesizer patch.  Which it is; my Sound Assistant has a tiny Ozone keyboard by him, and he performs the wind effect each night.

This last weekend I opened and closed "Starmites."  Four-performance run.  In that case, once again, the majority of the cues were presented in linear order; they were programmed into QLab, run off a laptop, and the Stage Manager herself was pressing the "Go" button.

I had a second laptop at the FOH mixing position.  There, I had copies of several of the sounds, several background loops, a foley-type effect and an electric guitar patch.

Taking these in order; the tech was abbreviated and the cast sometimes uncertain of their actions.  So there were a couple of sound effects I had duplicated so if the kids jumped a scene, or we'd messed up and forgotten to put in a sound, I could fire it from my keyboard instead.  The nature of these sounds (mostly magical attacks) is such that it wouldn't be a big problem if we accidentally both played the sound at the same time.

Which in fact did happen -- but as always I had thrown master faders for the sound effects on to the top layer of the mixing desk, so it was an easy matter to fade out the duplicate manually.

The background ambiance cues were on my computer because of the harried tech.  Even though I had the sounds built, it was simply too much to add them to the Stage Manager's book during what were already difficult technical moments.  These were low-level looping background effects anyhow so it was fine to just add them in to taste from the keyboard with one hand whilst I mixed the show with the other.

The guitar was there because the MacGuffin of the show, "The Cruelty," is basically an evil electric guitar.  We were hoping the band would do some guitar stuff as the prop was revealed, but that never quite I did some random fumbling live on a nice crunchy patch with a lot of echo (and a ton of bending).

The foley....first time I remember doing something like this was for "Honk!" where there was a whole bit about a man with squeaky rubber boots.  So I threw boot squeaks on to two keys of my sampler and followed the actor as he walked around.  

This was a similar gag.

So far, however, the only actor-triggered effects I have had were the Duck's "Universal Remote Control" for "Click Clack Moo," and a pistol used in a production of "Tis a Pity..."   The latter used a 424 kHz radio link to trigger a QLab sound cue.  The former was using a quick Processing sketch to interpret an XBee signal and play back a sound.

Oh, and an intercom buzzer for "Moonlight and Magnolias."  That was a strange compound cue; the practical switch on the prop intercom was detected by an Arduino that spit a MIDI message all the way upstairs to where Sound Cue was running on a PC.  Usually, the secretary was on "the other end" which was an actress backstage on a microphone that was fed into the same speaker.  But at one point she "connects" several other callers, which were pre-recorded voice-over sessions -- and these were played back over Sound Cue as Stage Manager "called" cues.

I've now worked on that signal chain so I have now a custom Processing ap that reacts to various inputs from battery-powered XBee radios and spits out a MIDI signal that can be picked up by QLab or by a sampler or Max patch.  I've used it with a Staples "Easy" Button modified for XBee wireless link, and with a basic accelerometer setup on a wrist band...but neither has yet been in a show.

Next on my programming chores is to add the ability to select a sound file for playback from within the Processing ap, to allow skipping the MIDI step entirely for simple shows.  But at the moment, that is my state-of-the-art in non-linear sound.

Monday, August 5, 2013


Most of my theater work is basically mono.

There's no point to panning the wireless mics attached to actors.  Even if you could track them across stage, the very reason for them to be there is so audience members sitting further away from the actor can still hear.  Panning the reinforced sound towards where the actor really is would be antithetical to that goal.  If anything you'd want to pan the other way!

More subtly, panning the band is sub-optimal.  Reason being that in many houses, a significant part of the audience is covered by only one speaker.  If you were to pan the second woodwind hard left, for instance, half the audience would hear too much flute in the final mix, and the other half would hear too little.

You can pan a little, depending on the house, and this can help to open up a congested mix.   But it is something best done in moderation, with an understanding of the compromises that may result.

In relation to that, one trick for getting vocal clarity (particularly important and difficult in underscored dialog) is to send the wireless reinforcement of the actors to one set of speakers, and the band reinforcement (or for some shows, the pre-recorded backing tracks) to another.

This leverages the Cocktail Party Effect: the way the human brain is able to sort out one specific voice from the surrounding noise...if given localization cues.

In smaller houses, the direct acoustic energy off the stage is your friend here.  If the first impulse that arrives at the audience member's ears is from the actor on stage, this helps their brain sort out the dialog they are trying to follow from the surrounding energy of band, ensemble, sound effects, and general stage noise.

Making this even more useful is what is called the Precedence Effect or sometimes the Haas Effect.  This is another aspect of the built-in human signal discrimination; localization occurs upon the first impulse if it is within a few dB (up to -10 dB in fact!) of a stronger but later signal.

What this means, is if your reinforcement signal chain is delayed by 5-15 milliseconds, the audience members will perceive the sound as coming from the actors, not from the speakers.  Even if the speakers are louder than the actors!

Tangentially related to this is the Broadway anti-flanging trick of sending each actor in a duet to a different signal chain; the so-called "A-B" reinforcement method.  This reduces the disturbing sound that arises when an actor is getting picked up on two different microphones at the same time (a situation that arises far too often in the romantic plotlines of the golden-age musical).

However.  However.  Of late I've been doing a number of musicals that are  more akin to the Brill Building than to Broadway.

This takes a little explanation.  For a classic like "Oaklahoma!" the illusion is that we are out in the wide-open prairie with Curly, Laurie and Judd.  Orchestral music swells in the background, and the people sing to us from where they stand on stage.

For a musical like "The Wiz" the illusion is of a Motown studio session.  Loud, compressed, artificial sound, band and effects and wireless all rolled together in to a single seamless mono mix.

The main advantage this gives me is that I don't have to try to taper my reinforcement speakers to try to mimic the inverse-square falloff of the acoustic energy from the stage.  Instead I am running a hot system that blasts right over the stage, and is essentially flat from the front seats to the rear of the theater; every seat, therefor, hears the same show.

It also means the whereas for a show like "Sound of Music" I took every chance to back off the reinforcement into naturalistic, near-invisibility, for this show the dialog is all run hot as well.  There is less of a division between song level and dialog level.

(However, this still doesn't release the FOH mixer from being conscious of the tendency for levels to rise.  You have to watch yourself, watch the meters, and consciously back down at intervals lest you end up having the system peaking well before the 10:00 number.)

(This bears expansion.  When you run hot, hearing fatigue sets in.  It sets in on you, on the audience, and even on the band and actors.  Even if your system could keep going up and up indefinitely -- if you had unlimited headroom and there was no such thing as feedback -- you rapidly reach a point of not just diminishing but paradoxical returns.  With every increase in volume, hearing fatigue increases faster and the sound becomes perceptually softer and more dull-sounding.  This process continues to accelerate until the audience is temporarily deaf and nothing you can do will help them hear the nuances of the next piece of dialog.  So watch those meters, check your own assumptions frequently, and take every chance offered by a slow scene or a soft song to back down again.  And then you will have the reserve power for those show-stopper numbers.)

I'm also, for the first time in this building, running without delay.

Usually I have up to 40 milliseconds delay on the reinforcement, leveraging the Precedence Effect to make the actors on stage the apparent source of all the audio energy.  In this case, I have no delay other than the physical distance (and the corrective delay between the different speakers of the multi-speaker setup).  The reinforced sound is the first thing to hit the audience, and all the noise from the stage and all the leakage from the pit orchestra follows in time.

And that I think has given me an unusual clarity for this house.  By presenting the reinforced sound to ears first, they are able to get crisp, time-aligned waveforms.  The smear of sound propagated acoustically from the stage falls behind that first peak.

As part of this philosophy, I've been doing a lot of effects designs that are similarly presentational.

In a realistic effects design, we pan towards where the sound is supposed to be occurring.  Actually, this is just the start.  We set up specific and unique speakers for effects.  We position speakers carefully to allow their sound to "Worldize" into their environment (a toilet flush sounds best, for instance, if you put the speaker in the part of the set that is standing in for a bathroom, and let the sound echo in a natural way around the set walls).

I've gone further, using bell ringers to make an actual phone ring on stage, putting speakers into an intercom box, even once using a live walkie-talkie on stage.

For a presentational effect design, the overarching concept is that what we hear as an audience is non-diagetic.

(The story is told by I think Alex North, who was asked by Alfred Hitchcock during the filming of "Lifeboat" where the orchestra was supposed to be in these scenes out in the middle of an ocean in a lifeboat away from everywhere.  The composer asked the director, "Where are the cameras?")

Anyhow, look no further than film or television comedy for non-diagetic sounds.  We don't suppose the "Waa waaa" trombone is actually in the scene!  (At least not mostly -- there are a number of classic comic moment when what was assumed to be non-diagetic turns out to be quite "real" -- at least to within the fluid reality of a comedy.)

So you conceive of sounds that reinforce the action or the emotion but aren't pretending to be the actual sounds of some actual thing on stage.  But even when the sounds are natural to the environment being presented, I have been tending in many of my recent designs to present them as part of the total mix going out the speakers.  Instead of sending an effect to stage speakers and pretending there is a pirate ship on stage, I send to the mains along with the orchestra and pretend we are watching a show that is presenting the idea of a pirate ship.

Sounds subtle, I know!