Monday, February 27, 2012

Easy....and crazy enough it just might work

I took a little time in the middle of microphone element repairs to do a Proof-of-Concept for the Easy Button.

Took apart the Staples Easy Button, de-soldered all the components and cut the traces on the PC board, leaving just the button. Ran a pair of wires out and hot-glued them in place so they wouldn't interfere. The button as built makes a lovely stapler "ka-chunk!" noise when pressed, but that was, sadly, inappropriate for use in an FOH station exposed to the audience. However, a little careful re-shaping of the metal spring took the "click" out without changing the firm feel of the button.

I connected it to my all-purpose Arduino-based wire-to-MIDI interface, and plugged that into QLab. The Stage Manager admired it so much she's going to try using it as a GO button for next weekend's performances.

Perhaps by then I will have gotten into my USB-AVR breakout board from Adafruit (or the old and sadly no longer made or supported Bumble-B) and made a more self-contained solution. Or perhaps not! This show is all about exposed technical bits. It's an Atom-Punk setting, after all. I've got a bucket of water with EL wire wrapped around it and a microphone sticking inside...I slosh the water around for a live sound effect in a couple of moments of the show.

(This is very much an insane cludge of a sound design. The effects are on laptop playing both through the headphone jack and through an Ozone USB keyboard for an additional two outputs. OS 10.5 was being stupid about combining these as an Aggregate Device but that turned out to be for the best anyhow. The drums are being submixed via a Presonus FP-10 into Cubase -- for EQ and compression. So I've got a maze of wires running all over the place.)

I've blogged before about the problems I've had in that space, and others, about the conflict between a flat coverage by house speakers, and an equal-loudness mix between vocals and orchestra. The fight is always against monitor leakage (especially when the pit is using amplified instruments like keyboards, basses, and guitars); since monitor leakage models as a large point source it falls off by roughly inverse-square. Since the vocal reinforcement includes delay speakers to cover the rear of the house, the fall off is whatever we set it for. This means that getting decent levels at the rear of the house is essentially at cross-purposes with having the same mix between vocals and band in all parts of the house.

Worse yet, the vocals are coming from distinct point sources high in the air (the center cluster in many shows) and the band source is a large, diffuse source below the level of the stage. This means the singers and the band don't sound like they are in the same universe.

This can actually be a help when you are trying to help the listener/audience discriminate between competing sounds. But the spacialization you really want is that the vocals appear to be coming from the actors....not from some speakers far overhead.

For several shows we've struggled with the problem of poor front fill. With the orchestra pit "in" (aka the elevator lowered and a moat between audience and actors) the front rows are just barely within the field of the main speakers. By ghosting up a center cluster or some high overhead speakers pointing nearly straight down, you can help those first few rows while not completely changing what the rest of the audience hears.

With the pit covered the problem gets worse. And when they chose to add a false proscenium (cutting off all your speaker locations) the problem becomes critical.

After a somewhat rough opening night and several complaints, I came in (exhausted and with less than thirty minutes to work) with a crazy idea. Front fill at stage level will solve this coverage problem of the first rows of seating, but many set designers will not permit the visual look of those speakers intruding upon the lower few inches of sight-lines. Stage fill from the front -- to allow the actors to hear the band -- can usually be snuck in at barely above stage level. I had these, in fact. Which were part of my monitor leakage problem already.

So what I did -- with no time to work -- is turn the front fill monitors around and re-assign them to the reinforcement bus. Since they couldn't of course be on the stage, they went to either side and were angled sharply in (focus point about a third of the way back in the house). And since for various reasons my subwoofers were already sitting on the far corners of the apron, I was able to stick them on top and lift up the speakers by that much.

Then I pulled both the mains and the center cluster from the house mix, leaving it ONLY the newly re-purposed front fills and the delay speakers further back. As the last step, I moved in the side fill monitors and re-angled them to make them the sole stage monitor coverage (the band is really, really loud anyhow).

And it worked. I was shocked at how good that sounded. But then, these were the original top fills and were a pair of lovely, lovely Meyer UPM-1P's. So they are entirely up to the task of being front audience fills.

I suppose if I was really concerned I'd re-focus the center cluster to boost a little through the middle of the audience (right down the center aisle is basically worst coverage from all the systems). I am also slightly concerned about audience in the seats closest to the UPM's, but there is sufficient direct audio pressure from the cast I think that will compensate. I have the UPM's at 8 milliseconds of delay, a compromise between the length of their throw to center and the average 12 millis the entire system is set to (the delay bias here is to allow the actors to be perceived as primary sources via the Haas Effect).

The rear house speakers are still on their Meyer-set delay via the Galileo processor. I did a quick-and-dirty impulse check using a side-stick sample via Garritan Personal Orchestra, and I didn't notice the sound doubling as I walked the speakers from the center to under the rear speakers. And I'm running the rear hot; as close as I've ever been to a flat field across the whole house.

Tuesday, February 14, 2012

Testing, Testing

I enjoy language. So here are a few more bits and phrases I've found useful -- these ones on the subject of testing.

Smoke Test: This is when you turn on the power and see if it smokes and catches fire. If you are really unsure if you configured things right (such as, perhaps you reversed the polarity on the power leads) the more refined trick is to turn it on for just a moment, turn it off before the magic smoke can leak out, then gingerly touch the most sensitive components with a bare finger to see if they are heating up.

A smoke test won't tell if you if the circuit works. But it will tell you if the circuit is broken. But please remember to always mount a scratch monkey!

Sanity Test: more commonly in the form of a sanity check (and I'm sorry to say, but there ain't no sanity clause), this is any sort of trivial throughput check or checksum or order of magnitude calculation; instead of the painstaking check of every line of code or engineering calculation, this is taking the simplest and most obvious check that will reveal that all your elegant math foundered when you multiplied instead of dividing in step 2.

This is why smart DIYers don't spare the blinkenlights. A few LEDs (or a few serial.print comments in a code) can tell you that what you intended certain parts of the circuit to do, they are actually doing.

In sound, a sanity test is done by ignoring all the nice microphones and nifty speaker processors and all of that, and just seeing if you can get a simple CD to play through the house speakers. If you can't, you shouldn't be wasting time setting up delay chains just yet!

Plugs-Out Test: I first ran into this phrase in regards to a terrible accident in space history. But let that not stop us from the idea of removing the umbilicals, and seeing if the device will still run on its own internal battery, without the connection to the ISP, and with the cover of the enclosure screwed down.

Proof-of-Concept: Not usually called a "test," this is similar often misunderstood by those who observe them. The proof-of-concept is done entirely to prove the plausibility of the idea -- it is in no way a test of the actual hardware. Indeed, you substitute, you breadboard, you mock up; whatever you have to do to get the thing to work, even for just a second or two before the tape falls apart.

Solderless breadboards, alligator clips, double-stick tape are your friends here. Simulated signals, simulated outputs. It doesn't matter what it looks like; it matters that it works, even if it only works once.

Coffee Test: This is the test of whether you are awake enough to write code or operate machinery; can you make coffee without doing something mind-bogglingly stupid? I grind the beans, brew hot water in a teapot, and pour it through a gold filter to the mug or flask for the day. This provides plenty of chances to prove I'm not awake yet. Today, I put on the water, cleaned the filter, and ground the beans. I finished just as the water began to boil so I quickly rinsed out the travel mug...then proceeded to carefully fill it with boiling water.

Other times, I've skipped the filter or the coffee entirely. The best day of all, though, was when I was still living in the Haight-Ashbury. I carefully filled a clean mug with milk, then proceeded to add coffee grounds to it. I remember watching the black flecks slowly sink in the full mug of milk and wondering what exactly was wrong with that picture.

 On a more serious note, there are a couple more test concepts that are worth adding.

Test-in-Place:  Especially with RF gear, funny things happen in the actual location where you mean to use the thing.  So it isn't properly tested until it is in the room it will be used in.  Better yet, it should be in rehearsal, or in as close to performance conditions as you can manage.

Acid Test:  This is when you create a worst-case scenario and see if the thing survives it.

Test to Destruction:  Unlike the above, this is when you intentionally fail the unit to find out just what it takes to break it.

The Big Easy

I am still throwing around ideas for dedicated Qlab control surfaces, music performance controllers, and linkage systems to allow sensor-driven sound effects, cue-driven servo events, wireless connectivity, etc.

Many of these options currently exist commercially (or semi-commercially, as in open-source kits), but they generally have two flaws; expensive, and fiddly. I'm thinking about robust dedicated systems; things with few visible controls and sealed enclosures that can be handed off to Stage Managers or percussion players or actors.

The trouble with figuring out the functions, and the form factors, of said things is the applications themselves haven't been adequately explored. There are so few options now for controlling sound effect playback from the orchestra pit, or linking a door opening sound to the physical door, or triggering a mechanical effect from a software cue package, that experiments aren't happening.

The design team is unaware of the tools the digital revolution makes available, and before someone like me gets a chance to say "I've got an Arduino-based gadget that can do that," the effect has either been semi-solved with traditional methods, or abandoned as being too difficult.

By the time you get up there and tell them the clock on the wall could easily point at a specific hour via a servo triggered by the lighting cue that is already taking place in that scene, everyone has become so adjusted to the idea that there is no practical way for the clock hands to move they reject the idea as being unnecessary.

I not too long ago went through a hellish sequence of trying to coordinate a slide projector, lighting cues, and sound cues -- when it would have been a one-day hack to control the slides from either of the previous, using DMX or MIDI or even assigning the "advance" button to a non-dim circuit.

In the long run, what I want is plug-and-play; to define a narrow subset of functions and make a device that just does that. One of the tricks to achieving that is making these devices cheap. I can't waste a full Arduino or worse on a single button. I need to work with a bare-bones purpose-designed board with no more than a cheaper AVR on it.

But until the point that tasks have been identified and the use of a digital system tried out in production, what I need instead is more of the experimental breadboards. Or, rather, the next intermediate; things with fewer dangling wires and software peccadilloes than what I have now. Yet, at the same time, devices that have more and more test bed functions on them.

Xbee networks. High-power switches and motor drivers. Sensors.

Which puts these devices in the horrible position of being development platforms in nice boxes with bland less-technically focused end-users can use them and not get scared.

Still, as part of the development I just need to have more building blocks already worked out. I need experience with Xbee links. With sensor conditioning. With motor drivers and high-power switches.

At this point I've generated only two specific applications, and they are so different there hardly seems to be a way to integrate them cleanly. One is the gunshot transmitter. The other is the dedicated Qlab controller.

The former is fairly easy to parameterize. The idea is that prop guns don't normally make a sound, and blank-firing conversions or starter pistols have a host of difficulties that make them less optimal for many productions. My test case was this; I gave the actor a key-fob radio transmitter, and he hid that in one hand, pressing the transmit button simultaneous to squeezing the trigger and jerking back on the prop gun.

The form I imagine is something that would clip easily and non-destructively to almost any prop firearm and detect the motion of the actual trigger via opto-interruptor or similar. Then a transmitter with a compact form factor would send a message out to whatever sound playback software is in use.

Possibly a better form is a butt squeeze -- this could be used on props that didn't have a functioning trigger -- but that would take more calibration.

Making a transmitter small enough to clip to the gun itself without it being visually distracting would not be easy, however. So you'd want a beltpack type transmitter, and a wire running up the arm. And then you have the problem of connecting...

Actually, I just had an even better idea. The "Trigger Finger." Use a bit of flex sensor in a flesh-colored finger cot. Run the wire under the sleeve back to a belt pack. In use, all the actor does is curl his finger sharply as if pulling a trigger. The sensor conditioner detects this and fires off the signal via the wireless connection. Aka it is a one-finger, no wires version of a data glove.

Because of the wireless link this is unlikely to be a cheap system. Of course, if one had spare wireless microphone packs around one could re-purpose one by sending a DTMF-type tone through it. But I find it is simpler to stay in the digital domain. A 424 MHz radio link gives you only enough range to get to a receiver planted on stage, but it is cheap. If you built the sensor conditioner and radio controller around one of the smaller AVRs, you could have a fab house run off a through-hole board that was barely the footprint of a 9V battery.

Alternatively, an Xbee or similar, though more expensive, can be eventually part of a complete Xbee sensor network. But as long as you are putting fifty bucks of computing and radio hardware in the belt pack, plus the cost of the housing itself, you'd best add battery charging and management as well, plus system monitors. Which brings us up to a multiple-unit kit cost in excess of $100 each.

So...I was wrong. The application is well parameterized, but the solution has not been properly designed yet. And even in prototype form, way too much fun not to try to have built by the next Makers Fair.

(Actually, the show I've got coming up includes a chainsaw that won't be practical but should sound like it works. In some other universe I'd be playing around with Kinect or Wii remote ideas to make the sound follow the prop...but for the flow of this show, a canned sound effect will be just fine.)

(I almost forgot an alternate strategy. Instead of going wireless, go light. Basically make a laser tag device, only with a bigger fan-shaped beam. If you want to be really clever, squirt a coded series of pulses like a TV remote does. Otherwise just tune your sensors. The cute idea...although it is of little use in the highly-rehearsed world of traditional that the gun can trigger multiple sound and/or practical effects by aiming at them in turn. It is possible existing laser sights could be re-purposed for this.)

((And as long as you are being silly, an infrared laser finger could serve -- so could a Wii remote -- to allow a person to play Tim the Sorcerer and trigger effects as he or she pointed at them)).

Anyhow. One of my flaky-pastry-item-in-the-stratosphere desires has been to wire up a Sonic Screwdriver prop with useful functions: Radio link to fire sound effects without having to go back to the sound board (very useful for a quick listen-through, checking stage monitor levels, troubleshooting, etc.) Laser pointer for explaining to crew which speaker is which, and also for use in rough-aligning speakers (I use a crazy device I call the tri-laser now; two laser heads glued to a carpenter's framing square). Emergency flashlight, which in the nature of such things is likely to be used far too often and run down the batteries all the time. Oh, and I'd love to have an SPL meter, a polarity checker...but by this point the device would be considerably larger than a tricorder and that's the wrong TV show entirely.

The other actual parameterized idea is the dedicated QLab controller. There are twin aspects to the application; the first is that while the "Go" button works fine, jumping or repeating cues requires hunting around on the keyboard, and editing cues puts the "Go" function out of focus. MIDI is always "focused"; even if you have hidden the QLab application and are working in a different one altogether, a MIDI "Go" command will work.

The second is that often booth or sound board space is limited. You can't always put a laptop where you can reach it easily. I'm actually quite fond myself of using a keyboard -- usually a battered old Korg Nanokey -- to control Qlab without having to have the laptop within easy fingertip reach. But I think there is a bonus, especially for less technical people, in having a small-footprint controller with big hardy buttons labeled with the universal standard "tape deck" symbology.

Actually, I just thought of an answer to a caveat I've had...the problem with remote use is you really, really want to see the computer's screen and know you are about to fire the right cue. However, nothing says you couldn't hack in a simple character LCD display showing "next" and "playing" cues. Since Qlab doesn't normally care about that, you'd have to leverage something like the MIDI or DMX functions and spend the time to add and write out special "invisible" cues that would send messages back to the remote console. So not really that elegant a hack!

A more complicated but satisfying routine might be to install a full graphic LCD...but then you'd have to both get that to be recognized by the laptop's video out, and drag the right items on to the resulting window.


Today, I am going to stop by Staple and purchase one (or more) of their "Easy" buttons. This will be my next proof of concept. I have already a few options in arcade buttons and the like, but I'd like to try wiring up this as a big red "Go" button. Or a "fire the sound" button for orchestra pit.

In the development form, just use an Arduino to spit MIDI or MSC. Or use my AVR-USB development board to spit out an HID-standard "space bar."

In the next ranging, figure out how to do MIDI on a naked AVR, or even bit-bang MIDI on one of the ATtiny's with no UART (which is really just an exercise in clever...there is no good reason not to just use an ATtiny45 instead). And then, figure out how to do MIDI-over-USB....HID is relatively easy (or so I am told!) but MIDI-over-USB is a tougher trick.

The ultimate controller form would be USB-powered and USB-linked, with some degree of feedback (at least lighted buttons) and a minimal display to allow entering program mode and changing the system defaults (such as, switching to MSC protocol).

Anyhow. Swartzbrot awaits. I'm off to run errands.

Monday, February 13, 2012

Quick and Dirty... a bunny with, err, dirt on it.

My horrible tech is finally over and the show is open. We're not working there again. Not unless they make some major systemic changes.

You can fake your way through some shows with a "turn the mics on when they sing" approach but you can't fake your way through a loud, pop, in-your-face modern musical like "Wicked" or "Rent" without an actual FOH mixer. Or something very clever to replace them with. We tried. We gave up. I dismantled most of my stone knives and bearskins. Un-taped the twenty-dollar microphone from the chair and stripped out all the orchestra reinforcement. Rented some new units and put the wireless transmitters that are older than some of my cast members back on the shelf.

I've got the entire vocal bus compressed at 2:1 and with a 3dB presence peak thrown on it as well. Horrible, horrible things to be doing to the poor sound. We stuck half the cast on madonna-mics to get the very last bit of gain before feedback. And we told the client we would never design a show for them again.

On a lighter note, I made my own microphone. Sort of. I was expecting to record some location sounds for an upcoming production and I wanted a way to use my $20 boom pole with the minidisc recorder I purchased ten years ago in Tokyo.

Well, I've been repairing wireless microphone elements for the last couple of weeks, and I had several Shure WL185's lying around on the desk. These are giant soup cans of an element and hardly worth sticking a new connector on them as they are a bit large and clunky to tape to an actor's face. (I do like them fine for interview-style lapel mics, though.)

So I took two of them, zip-tied them to a simple coat-hanger support, and spliced them into a single TRS mini-jack. Since they are electret condensers, my Sony MZ-R900 has no trouble powering them. Handling noise is pretty bad but the sound isn't bad once they've been stuck in a mic stand.

Next experiment is to make a psuedo-MS mic out of three elements (two back-to-back cardiods simulating the figure-8, and the third for the front element). Or, since you can get a similar element from Digikey (or a cheap knock-off from Radio Shack) -- the naked element without the nice housing and grill, that is -- I may go the whole DIY microphone thing and build an Mid-Side from scratch.

Why do I want a Mid-Side, specifically? Because that allows me control of the stereo width/room tone. For recording sound effects on location, this is a simple way of having some control over the apparent distance and presence. It just requires doing an M-S matrix during the mix-down.

Or, alternatively, I could rig them in an ORTF configuration. I set them up as coincident pair because that takes less space...ORTF is technically 17 cm in width. Coincident pair can also be summed to mono without phase cancellation, but I'm not exactly worried about that!

I find myself thinking of clever mechanisms that could spread out the elements on little pivoting booms, but such things tend to, by the time you are finished tinkering, cost more than just running down to the pro audio shop and purchasing a pre-made equivalent would be.

The next show has only four pieces in the orchestra pit, but it is a covered pit and the actors are stomping on top of it (all of the dance numbers seem to be really, really stompy!) So this is actually a good reason not to either leave the drums alone, or make do with an overhead and maybe a kick. I'm going to actually go rock and roll and throw as many mics as I have channels at the poor percussionist. Because with all that stomping, I have to get pretty close and isolated with everything down there.

But it is still going to be quick and dirty.

Friday, February 10, 2012

Stone Knives and Bearskins

Two highly technical shows opening within a week of each other. The first opens tonight.

800-seat house, nearly sold out for opening weekend at $45 bucks or more a seat (not counting that thrice-damned GoldStar), 14-piece pit orchestra, thirty-four cast members one of whom was the star of the Broadway production. And, oh yeah, the half the cast are singing into elements salvaged from the recycle box, the band is going through an old Mackie mixer and on the bass is a twenty-dollar mic gaff-taped to a chair.

Show business is of course all about wood spray-painted to look like gold, canvass walls standing in for marble halls, and princely robes salvaged from old fur coats. But this is the most absurdly old, damaged, duct-taped together, barely tolerable gear I've ever been forced to apply to a show that is intended to sound this tight, slick, pop, and in-your-face.

The quote I've been using all week is from original-era, the-one-and-only "Star Trek." Spock is trapped in the past, and is trying to build an interface to his tricorder with what he can afford to get at a ham radio supply store in Chicago in the middle of the Great Depression. A local walks in on him as he is in the middle of struggling with his jerry-rig, and he snarls, "I am endeavoring, ma'am, to construct a duotronic memory unit utilizing stone knives and bearskins."

My wireless mics are a grab-bag of different brands, over half of which lack the punch to even make it to the sound booth. I have receiver packs tucked into every corner of the stage, with miles of XLR running every which way trying to get to and make use of every last one of the sadly limited house wiring. One single solitary snake would help immensely, but the over-protective house goes prompt critical if anything like that is even suggested.

My hookup is a nightmare, and another nightmare faces my A2 as he struggles with multiple mic changes. We've changed so many things in patching around problems there is no longer anything resembling an order to who gets what and where it is located and what input channel it uses. And the directorial team keeps asking for additional voices to be added to the overstressed list.

There is no frequency plot and never will be. Hetrodyne interference is just a fact we live with.

I'm spending hours every day with fine-blade x-acto knife, heat-shrink tubing, and a magnifying face shield repairing breaks in the tiny 2mm cables and rewiring the tiny TA4 and locking 3mm mini jacks and trying to turn more of the bucket of tangled old dead elements into something we can use on stage.

The pit is ankle-deep in cords (which is not unusual for a pit, but I prefer to keep my pits much neater). And as scary as the mix of old cranky underpowered microphones on stage is, the fact that the band is sub-mixed on a mixer that no-one is watching is even scarier. The only control we have from front of house is to turn the submix up or down.

(Well, okay, at this point in the evolution of the hookup, I could break it into three sub groups. But I am not even sure I have the channels left on the FOH mixer to handle that, the sound booth is badly designed and you can't hear what you are doing from inside it (!!) and the mixer has enough on his hands already without trying to mix a band as well.)

So I spent preview night mostly standing behind the brass section, setting up a band mix on headphones. Tonight we hope to try it....ONCE!! the house speakers before we open the doors and let the opening night audience in.

No pressure.

But on the whole, I'd rather be knapping flint. Bearskins I can do without, but stone knives actually sounds kinda cool.

Thursday, February 2, 2012


At pretty much every level of this business, if you want the best performance you will be bringing your own gear. The difference being that at the classier places you bring gear because you are used to your own stuff and can work more efficiently with it; at the mid-range places you bring your own gear because the house doesn't have quite the same quality in all categories; and at the lowest levels you bring your own gear because the house doesn't have any at all.

Pretty much goes with this equation that this is a cost out of pocket. Few houses will reimburse for the use, or the ordinary wear and tear, of all the mics and headphones you bring. (I bring my own CABLE most places -- because the house cable is rarely stowed correctly and doesn't get tested quite so often, and load-in is too short and too hectic to deal with untangling some-one else's mess only to end up with the crackle of a bad connection.)

There is a bit of horse-trading going on too, but the problem is, from the middle on down everyone is shy of horses. You might have an extra cart to trade at one place, and extra hay at another, but EVERYONE needs that extra horse and no-one has one to spare.

And that's where I am right now. Two of the most complex shows of the season opening simultaneously in different venues. Where normally I'd have a bunch of powered monitors, a small selection of mics and stands, my venerable Mackie mixer, Lexie processor, and of course multiple channels of wireless mics to dedicate to one, this month I am facing a situation where two (or, rather, THREE) different venues all need the stuff.

And not a one of the venues can afford to pay rent for the gear they'd normally get from me for free. And of course since all three productions are simultaneous, none of these places can afford to lend to another one, either.

So I'm working through paperwork now trying to stretch four circuits to cover eight instruments, 19 wireless mics to cover a cast of 32...and also looking at trying to use some old Yamaha speakers as floor monitors (to replace the nice monitors I'll be using in a pit elsewhere), and in true trickle-down fashion, finding a really, really big hiding place to get the old house speakers re-installed with whatever amp I can scrounge so as to replace the missing effects speakers...which in turn are going to try to be a half-assed front fill.

Well, back to the paperwork. I've been at it for nearly twenty hours now, and have gone through six different attempts to graphically abstract the data so as to detect possible patterns. I think, seriously, before the next time this comes up I'm going to take a little time to crank out a software program to do this for me!