Audio Systems are Computers

TLdr; Settle in, it’s a long one. I didn’t want to break the flow into two parts. Not a comprehensive history but one from my perspective and we were in the middle of it. We go from amp monitoring to system measurement and control in the early 90s to full blown programmable system drives in use today. It’s been a long road because of the development of the technology and the fact that not everyone wants to interoperate with everyone. Once again in terms of implementing cutting edge tech the industry is its own worst enemy.

It’s been just over 30 years since computers started to be married to audio systems. These days audio systems are computers. Even if the components are connected with analog interfaces the devices are microprocessor based. Except for some vintage or boutique gear nearly everything you use manipulates the signal digitally. Increasingly it’s also transported digitally end to end.

Thanks for reading A Barking Dog! Subscribe for free to receive new posts and support my work.

My first exposure to interfacing computers and audio systems came in 1990. I just started with a company up in Seattle. This was when Seattle was Windows, Super Mario Brothers and jets. The company had just completed a baseball season long lease in the Kingdome of an EV MT system controlled by TOA Saoris and powered with Crown Amps and monitored with something called an IQ control system. Based on the lease they got the contract for a complete Kingdome install as well as a similar install at the then new Alamodome. A Macintosh SE/30 was used to control the IQ system and an AcerNote laptop for the serial connection to the Saori though only one at a time.

I would also be given a mandate to migrate the flagship touring system to computer control. From that point I took every opportunity to try to converge computers and audio. Some successful. Others not. It took a couple of years but we finally managed to build one of two large touring systems on the road with remote amp control and Saori control. Though not specifically controlled by a computer we also added a tc1128 eq system and 6032 fader pack remote. The other big touring system with computer control was Stan Miller’s system he used with Neil Diamond. That was it at the time though I think Stanley built his a bit before we did. The Kings of Grunge and the Sweet Caroline guy.

The rush was on. Crest released the Nexsys amp control system and others were working on solutions. The most ambitious was a product called Visual Network Operating System by Lone Wolf Systems in Seattle by way of So Cal. It showed such promise early on that they were able to secure funding from Microsoft co-founder Paul Allen’s Vulcan Ventures. Soon they said you would only need one network and one computer to control every audio device in your rig. Real soon now. No, really.

The network was branded Medialink. Then, just like now, everyone used their own control protocol most of them a current loop or serial. Though not TCP/IP based Medialink was similar to today’s IP audio LANs. A proprietary chipset was to be developed with the software auto discovering any Medialink device.

I was able to use one in the prototype stage in production rehearsals for a couple of gals that had this local band in Seattle. They were sisters and sang about fish, dogs and butterflies. The VNOS rig didn’t quite work and that spooked the house guy so he didn't want to deploy it. Normal development teething pains. It’s not like it was going to shut off the system. I would have used in on the mons but it sat unused for the duration of the tour. A couple years later after not delivering the chipsets to licensees at scale Lone Wolf went out with a whimper and some bad blood from the licensees.

AES was quick on the case with a standard called SC-10 which was to be an open networking standard for audio device control. It never gained any traction in spite of many of us pushing it. Everyone wanted to roll their own. Meetings, panels, seminars, politicking, begging. No one wanted to play except in broad academic terms. No one wanted to be the first and the standard kept changing. As an industry we weren’t ready. Ethernet and IP weren’t widely in use in smaller business settings let alone entertainment technology. I can’t recall anyone seriously discussing it. In fact were were using either NetBEUI or AppleTalk in our business networks. We were looking at Novell’s NetWare for expansion of the business LANS rather than ethernet/IP.

Many of us had been controlling effects units from laptops using MIDI and Buford Jones’s excellent Hypermidi driven Hypercard stack. It was the first show control/cue reference I saw. There were also some librarians including the unofficial SPX90 librarian. A trusty PowerBook 140, a copy of Hypercard and an Opcode interface in the serial modem port and you were good to go.

About that point in the mid 90s active crossovers developed into full blown system controllers. In addition to crossover shelves there was equalization, delay and polarity control. Most were standalone with no external interface but that was about to change. Peavey released something that would change the direction of computer and audio integration. Known for low end musical instrument electronics they jumped feet first into commercial audio with the Media Matrix line. Not only was it computer controlled system but the system was the computer. Add in wall panels and touch controllers and you had an integrated page/AV/install system. Many, myself included, were dismissive that Peavey could have built such a thing that also sounded good.

That changed in ‘94 when I was prepping a touring in Balmer. I was working for a band that were unsung once too often. Could not rub off. As part of the deal for the insanely low price I was to prep my monitor rig. About quitting time I looked up and no one was around. The shop was massive. I followed the din of what sounded like pink noise then music. Somewhat like the band I was working for at times. I found two large clusters not quite arena sized. About 100 feet away was a console and a stack of Media Matrix. I had stumbled on the secret room where they were testing Media Matrix. I was sworn to secrecy at the time as they were building a touring control system with Media Matrix to employ on a tour with their client who at the time was the reining pop singer from gospel roots. The flexibility, quality and ability to manipulate aspects of the drive system were outstanding.

There was no turning back. It was all about digitally controlled analog or digitally manipulated signal with analog I/O. Everyone had one. We settled on the Omnidrive line from BSS. We also had KT DN3600s and BSS Varicurves. Some you could hook to a computer but there really wasn’t any utility touring since the front panel interfaces were full featured and easy to use. It was handy for installs though. As the years passed many others from BSS Soundweb at the time morphed into London and eventually Q-Sys showing up. Control systems got ever more complex with offerings from AMX and Crestron. As others came to market with limited or no front panel control the utility of a computer was no longer a question. It was a necessity.

The design changes not only reduced the prices but also increased the density and rack space requirements. For touring hat was a double edged sword in that many liked interfacing gear with a front panel interface with a computer but didn’t quite want to go full monty and leave the front panel interface behind for a computer only. For installs ditching the interface and locking in the settings was a no-brainer. This tech was still pretty new and some touring crew weren’t yet carrying personal computers let alone using them to control or monitor systems. It was going to take some time for people to adjust to it and for the price point to be reduced to where it was standard equipment. Installs took to it like a duck to water.

Touring attitudes toward computers and audio were about to change. In the mid 90s JBL released JBL-Smaart. It was a 2 ch FFT analyzer that ran on a laptop with an audio interface. It was licensed to the Orange Umbrella by SIA Software. SIA was Sam Berkow and Alexander “Thorny” Ulli Thornton II. Fast Fourier Transform and transfer function weren’t new. While the concept of FFT is from the 1800s legendary Grateful Dead audio wonk the late great “Dr” Don Pearson was experimenting in the 70s with using it for tuning audio systems. That eventually became Meyer SIM. It was pretty expensive and not widely used outside of the Dead or Meyer circles. What was in use particularly among acoustic designers was the TEF or Time/Energy/Frequency measurement.

A TEF utilizes a burst of tone to create the stimulus for the measurement then it’s analyzed when the measurement it complete. This sucks for our use because we can’t do it during the show and even prior to the show it’s damn annoying (though it’s arguable that playing the Nightfly when tuning is damn annoying as well). Using a transfer function was the ticket because you could use it during the gig with what you’re mixing as the reference. That and the pricing were the selling points. It was about $700 in the mid 90s which wasn’t cheap but compared to the alternatives it was a bargain.

This was an epic shift in pro audio. The angst that many had about hooking computers to systems was quelled as this was a measurement tool independent of the system. We needed good measurement because qualities like coherence and pattern control were new or nonexistent in many if not most cases. Systems were sound sprinklers when compared to the systems of today. This was the most impactful use of computers and sound systems to date.

QSC entered the amp control market sometime around there. For an installed application it’s a great feature as you can monitor your amps in real time and be alerted to any potential failure. For touring it was cool but wasn’t required for most acts. Now it’s normal and built in. Then it was an option to be purchased. QSControl was also the first ethernet/IP based monitoring system. Around that time the first implementation of digital audio over unshielded twisted pair cable was CobraNet. CobraNet was developed by Peak Audio and marketed by QSC as RAVE. I’d been at a new company for a while and we had changed to QSC from Crown. As part of our deal with QSC we picked up the hardware for RAVE.

CobraNet worked as expected. The fidelity wasn’t quite up to spec for high level concerts but was good for paging systems at the time. The selling point was that a butt load of cable and conduit could be eliminated or at at least minimized. It was reliable so we took it on tour with the guy from Portland with big black boots and an old suitcase. We used it primarily for pre soundcheck demos but did a few gigs with it. For most stops on that tour the RAVE system went unused. As they hammered the shit out of the PA the QSControl system was pretty much a must have. The fact we would WiFi into it was a plus (via one of the early AT&T WiFi systems available). No delicate UTP network cable to get trashed by the moshing punters. CobraNet is still in use today but has largely been supplanted by Dante.

Not much changed for the couple years with regards to computer controlled audio. Until the Yamaha PM1D came out. This was the turning point from analog consoles to digital surfaces. Many in the analog world didn’t take to having an all digital console. Now the console was a computer. It was just packaged like a console. It wasn’t the first programmable mixer but it was the first to be taken seriously by touring pros. Introduced in fall 1999 I didn’t see any in the wild until mid/late 2000 perhaps early 2001. It sounded like one would expect a Yamaha console to sound. Clean almost sterile compared to something like a Midas or Soundcraft that had a bit of a bite in the audio signature. Price was the hurdle on this one. Around that time there were other options for recallable control but those were limited to digitally controlled analog.

In 2002 Digico upped the standard with the D5Live. It had a sound more similar to that of a Midas than a Yamaha and featured a touchscreen interface. In spite of some early pains the D5 proved to be more popular at the time among those of us in the US touring set. By 2003 I was specing them as often as I could though on smaller tours and gigs price and availability was an issue. Around 2004 Yamaha struck again with the PM5D. The PM5D did for digital consoles what the PM3000 then PM4000 did for analog consoles to make Yamaha the regional and mid range touring standard. I’d been carrying a DM2000 mixing mons for the guy who figured out there was no dark side of the moon. It was an easy ask to swap for a PM5D for most of the gigs I was doing. Except those gals in Seattle that dug the sound of an XL4. At that point I was either Digico when I could or a PM5D if I couldn’t. PM5Ds became ubiquitous.

Almost overnight people started ditching their analog consoles. In summer 2005 I was prepping a tour in the west coast shop of one of the big dog PA companies and the shop was dead bare except for our rig and half a dozen unused Midas H3000s. Once the touring standard they were relegated to holding the break room door open at JD’s shop. It was literally holding the door open. Except for the most diehard old dogs, including one that still uses his PM4000s to this day, the age of the large frame analog console had ended. The flood of new blood came rapidly. The D Show, Profile and SC48 from the Avid camp. Everyone either had one or one in development. And just like that analog consoles were dead.

Speaker systems and configurations were developing swiftly as well. With Christian Heil’s further work on Harry Olson’s original line source coherency and pattern control were becoming the norm. More has happened in the area of sound reinforcement speaker array development in the last 10-15 years than in the history of live sound. While much of it revolves around horn and waveguide design a not so new application called beam steering is also at the forefront.

To use this new-old technology it needed calculations other than rigging loads to be able to deploy. We needed to know what angles to splay the boxes as well as the most optimal trim height. The first software efforts were on spreadsheets and fairly crude but they worked. That developed into 2D visual tools then into 3D visual dispersion and acoustic prediction tools used in todays touring systems.

When the modern line arrays were first deployed two of the leading manufacturers were dead set against using any array shading or different processing along the array. They were militant about it. While only one had a dedicated processor (bundled in the onboard amp) the other had a firm requirement for amps and processing including rack configurations. This was the start of large scale “black box” processing where they user has minimal or no ability to tune the configuration other than eq.

We weren’t working with those folks though. We had a fair amount of experience array shading and multi processing rows of point source arrays. We had good success using just shading, recalled and controlled from an amp control system. The results from shading only led us to branching into using processing along the length of the line to further enhance fidelity of the array. If it sounds good it is good.

Eventually most came to realize that with the right tools deployed in the proper way array processing had advantages. Modern methods are much more complex than simple shading and eq. To implement the tech there needs to be a base of engineering to support the configuration. You can’t just pull a config out of your ass. You can’t calculate it using a spreadsheet. That’s when processors with array optimization functions appeared. This has further morphed into fully programmable drive systems supported by acoustic simulation software using more advanced filtering like FIR and all pass filters. Now the computer is the system. End to end.

Modern drive systems require training and the ability to implement the tools to get the best result. Same is true for consoles as well. The use and knowledge of a computer is no longer a luxury. It’s a requirement. Add in network transport and monitoring and control apps and in some ways it’s more like an IT gig than an audio gig. I don’t see us being done with the tech curve just yet. As multichannel spacial systems are more widely deployed in touring and concert audio the configurations will be more complex. This will require more realtime processing with the ability to quickly and easily modify designs for different rooms.

The knowledge and skills for prediction and and tuning will be greater as the tasks grow more difficult. Tools like audio object based design make this more accessible but there is still complexity. In terms of creative representation of complex soundscapes we’ve yet to reach our potential. The best is yet to come. And with it complexity.

Thanks for reading A Barking Dog! Subscribe for free to receive new posts and support my work.