16 channel soundsystem in the 1-st fort of Kaunas
Imagine this – I call Marius to tell him that we need to tear down a greenhouse – a good opportunity for him to sample a lot of greenhouse. He responds with – “I can’t, I took a project where I need to install a multichannel soundsystem in a military fort.”
That’s one way to get me intrigued, so I let the greenhouse to rot by itself and proceeded to get into my car and go to Kaunas to see what’s going on.
The project was initiated by an organization known as “Fortų Įgula”, which is focused on renovating and repurposing the forts of Kaunas.
The objectives:
A multichannel amplifier setup.
A multichannel speaker setup.
A multichannel interface, allowing to playback multichannel sound.
Capable of 24/7 operation, preferably without a computer.
Robust enough and easy to maintain.
Budget and time available – aaaaaaaaaaaaaaaaaaaaaaaaa
As quickly and as cheaply as possible. Great.
Starting phase – coming up with an idea.
We initially planned to buy a lot of used 2.1 systems you’d find in grocery shops, tear them apart of their amplifiers, put all of them into a single box with one power supply and just use long speaker cables to connect them to the speakers.
The main concern was that these systems might not have enough BASS for the large areas in the fort, but we were unaware of how reverberant they actually were. I brought a pair of those systems to test and it was immediately apparent that loudness isn’t going to be a problem, thanks to the insane reverbaration.
And we definitely didn’t need any clarity – the environment is noisy by itself, sound would be coming from pretty far away, washed in reverb and for a budget that didn’t leave a choice.
And then of course we needed to know the conditions – the resiliency of the electrical grid, potential for ground loops, level of moisture in the air, length of cables, etc. so we’d know what kind of countermeasures we need to install into the contraption.
None of our concerns were true, except for the length of cables, which didn’t end up being an issue at all.
Architectural asymmetry.
One of the issues was that the fort itself wasn’t symmetrical, therefore we couldn’t maintain the symmetry for the multichannel system. We couldn’t adhere to any particular standard (like ambisonic or whatever).
We went with an alternative approach, by selecting several adherent rooms and instead of having each of the 16 speakers dedicated to a central point, we made six rooms with their reverberation acting as a source of sound instead and the speakers are dedicated to the given room, rather than a dedicated listening position.
The objective was to fill the area with sound, so we intentionally positioned some speakers facing away from the center, so we’d be able to “emulate” action taking place in a given room and how that would be perceived in a different room.
Any other approach would’ve not created any control over the perceived spaciality of the soundsystem, because we were working with more than 6 seconds of reverb – forget any directionality, or locating where the sound is coming from.
Cheap is too expensive.
First of all, I was confused with the idea of using the 2.1 systems, because they’d all be different, power ratings, frequency response and would take a while to not only modify, but source – I’d have to drive around two cities and they weren’t cheap enough for the budget.
I had to delve back into the classifieds for the lord saviours that are garbage hustlers. As I was looking for speakers to buy by weight, I couldn’t find anything cheap enough, already browsing in slight desperation. Until I browsed in “Music equipment” instead of “Speakers”.
And oh my, did I find a deal, I just needed to look in the wrong category.
There was a guy, who was selling some microphones, loads of other stuff and a 29 speakers consisting of Mach CL4 and CL6 installation speakers, meant to be used in arrays, to cover a large area.
And it was too cheap to believe, because they are water-proof, two way, with an actual crossover, made out of plastic. I made a call, immediately arranged a purchase and on the next day, I hopped into a car and bought all of them. I have remaining CL4’s to this day.
The amplifier:
This solution, created a problem – we needed to build the amplifiers. I didn’t have enough experience with building my own amplifiers, nor did I have enough time to, therefore I consulted a friend of mine who has experience in sound installations and opted for TPA3116D2, 2x50W (stereo) class D amplifier PCB’s. I bought ten of them, in order to have some spares.
We assembled the box out of scrap wood that I had, painted it black, made a plastic faceplate with terminal connections, soldered everything to the PCB’s directly and came up with all sorts of names for it, referring to the D jokes, since it housed Class D amplifiers.
The power supply:
We ended up using PC PSU’s, because they are cheap, reliable and easy to source.
However, I needed to remake them slightly, to make them suitable for powering a lot of amplifiers at once in parallel. So I bought two, using one PSU for one row of 4 amplifiers, to not overload the +12V rail of the PSU’s.
Inputs, outputs:
I routed the entire faceplate in this configuration:
You can input all 16 channels separately, or:
You can input 8 channels with speakers playing in pairs.
There are three jack inputs, two for stereo and one mono. The input can be switched via the tumbler.
Speaker terminals to connect the cables that go to the speakers.
A knob, that serves as power on and volume control.
The soundcard:
The biggest roadblock of this system, was the soundcard. We wanted to avoid the use of a computer as much as possible. Therefore we looked into every possible option.
Only one of them made the entire thing possible – Cymatic LP-16.
Otherwise – I would’ve bought 2 used MOTU 828’s or Ultralites and then chain them together to act as one (since I know that’s possible with an 896 and an Ultralite) and use a laptop.
The Cymatic LP16 could play back from an SD card through 16 unbalanced outputs. And it wasn’t that expensive. However:
It got cancelled and the company closed, right after we bought it from thomann.de and the unit that we got, was broken – one of the jacks broke off inside the unit, therefore we lost one of the outputs. We ended up utilizing the headphone output instead.
The sound art:
Marius is the one who composed the sound art for the exhibition. It lasted ~15 minutes and was done in the last four days remaining for the project, while the workforce installed the speakers in place.
He mostly used things recorded in the fort and/or played back through the system itself, in order to capture the sonic signature of the fort’s chambers.
Also, in order to create the feeling of movement and action taking place in the rooms, Marius programmed a massive Patcher, that allowed him to pan things around those 16 channels.
The exhibition:
After the sound art played through several times, the party was over and sadly, our efforts to make the system usable without a laptop, with cheap but robust Class-D amplification, with fans, waterproof speakers, were completely unnecessary.
There were no more exhibitions and the entire thing is no longer used. If I recall correctly – the speakers are used indoors in another fort.
It was a fun project, even though I didn’t know half of the things I had to do and I can’t be mad about it. I assembled the fastest available PC for music production out of the comission I got.
What’s next:
Since we still have access to the fort, we’re going to bring most of our microphones, build some transducers and portable speakers to record some absolutely sick reverb impulses.
The other part of the fort which we explored is a massive tunnel complex under the ground. It might sum up to a kilometre or more of tunnels of various widths connected together and it produces more than 15 seconds of reverb.
Here’s some hear – we found a rotten piece of OSB board and a plank. Therefore a bass drum. The recording is done with a phone, so it’s mono. But you’ll get the idea.
Subscriebs to not miss out on the impending impulse response pack.