I love engineering things — toys, robots and software — and building stuff with my friends. I also love learning and teaching, which is why I wanted to share a few details about our latest passion project, Bass Bar. Please enjoy and feel free to ask any questions in the comments.
My builder in crime is my friend Tim. We first met as engineers at Nitro and immediately found ways to collaborate at work as well as outside the office. Both of us are former DJs and our shared love of music and tinkering with both software and hardware meant that we were coming up with ideas faster than we could build them. We built a few things together over the years — Industrial robots, VR systems to control said robots, high altitude weather balloons, robot costumes, and lots of 3d printing. It was always fun but they still felt like weekend projects when we wanted to do something that really pushed us to grow.
We needed a stage and community that would afford us the freedom and creativity to build something extraordinary. Burning Man! A blank canvas for ideas guided by ten principles in very challenging conditions. Having been to Burning Man several times before, Tim and I thought of what we could build that would be challenging, fun, and somehow contribute to the community. We knew that we wanted to make music and doing that in the desert is hard but was an interesting thing to solve for (hint: more sound!). So we set off to make great sound in the desert and sparked the idea for Bass Bar.
It’s pronounced “Base” as in “Treble & Bass” and not like the fish…
Idea to Design
Bringing any idea to life is hard and we poured a lot of our professional and personal experience into making Bass Bar a reality.
Our past experience developing and deploying large scale applications and infrastructure gave us the ability to execute quickly and effectively. Coming from a CGI background, I was able to design concepts that guided us through the various constraints we would eventually encounter. Everything from managing its size so that we can transport it hundreds of miles to the desert, to its overall aesthetic and ability to withstand the elements. We also had fun visualizing what it would look like in the desert with some photorealistic like the one seen below.
Iterating through visuals enabled us to tangibly react to each other’s ideas, quickly finalize our design, and provided a foundation for us to build from. The design process also helped us organize our concepts into four key elements.
With a clear picture of what we wanted to build and how each component fits together, we were ready to start making stuff.
Design to Build
All your bass are belong to us
You can’t be called “Bass Bar” and not have bass. We didn’t want to simply add ten 18" woofers and be done. We wanted to balance and model our sound.
To create this, the challenge wasn’t just being louder, but also factoring in how sound physics work. Sound is a function of pressure — the more pressure you build, the louder sound becomes. This is especially challenging outdoors where there are no objects for sound to bounce off of.
We’ve definitely experienced great sound at home that sounds terrible at Burning Man. Think about this, why is there no sound in space? It’s a vacuum, which means there is no pressure (I know, I know, you also need a medium for the waves to propagate through, but this write up is about “Bass Bar” and not “Space Bar.” Not yet at least.).
All speakers work on the principle of physically pushing air particles. They do this by oscillating back and forth based on the input signal. Notice when waves from the rear are bounced inside a cabinet they travel further than the frontal waves.
Because of what we needed and our budget, off-the-shelf subwoofers weren’t really an option. While they add more bass, we would have spent thousands of dollars to get only marginally better sound. Our solution was to design our own passive speakers and purchase professional sound gear to model our sound. This consisted of power amps, crossovers, signal processors, and distribution amps.
Modularity was another major concern for us. Anyone who has built anything from APIs to car stereos understands the importance of keeping things modular. We were able to design all of our cabinets to function like LEGO bricks which gave us a physical modularity — move speakers around, add more later, upgrade existing, etc. The next modular piece was our sound modeling. We designed and built a club-level sound system that gave us enough flexibility to model and rearrange components to play for different scenarios.
So what does this setup look like?
- 8000 Watts of power (to power 30 Speakers)
- 2x 18” woofers
- 2x 12” woofers
- 6x 10” woofers
- 2x 8” woofers
- 8x 6” 2-way speakers
- 10x tweeters
- 2 Distribution Amps
- 2 Signal Processors
- 1 Crossover
Burning Man really lights up at night so we had to pair our sound system with equally absurd lights and visuals. We did this with several independent elements consisting of hundreds of LEDs, a custom designed “Bass Bar” sign, and an outlandishly large video wall.
Connecting hundreds of LEDs is easy nowadays. Several smart people have made invaluable libraries with countless documentation. Video walls, on the other hand, are very difficult to wire up and documentation is sparse so we had to figure out a lot by learning along the way.
Video walls are made up of dozens of LED panels that are wired up to make a large display. These panels range in density (1.5mm pitch vs 5mm pitch) and pixel count. We decided to order a palette of these panels from AliExpress in a P5 size, which was a good balance of pixel density and size.
Controlling the Wall
You can technically control a single panel with an Arduino, as we’ve explored through hackathons we’ve hosted.
The problem is controlling tens of thousands, if not millions, of pixels from a digital video source. You not only have to worry about powering all panels at once, you have to worry about timing and syncing. While our video wall is not very dense, (our resolution is closer to an original NES than an HD display), it still clocked in at 448 pixels wide by 192 pixels high, creating a whopping 86,016 pixels to control in total.
We had to use a controller board (we originally bought ours from Adafruit — bless Lady Ada and her amazing company) that can handle such panels, but they’re made in China and the software to configure is a combination of Chinese and Easter eggs (Seriously, the first app we used made us hit a special combination of keys to unlock the configuration screen).
In the end, and almost 100 hours later, we managed to configure, build, and power a beautiful display.
- 8’x4’ Total size
- 42 P5 Led Panels
- 448x192 pixels
- 1800 Watt max
Lastly, we designed a custom Bass Bar sign that would act as a beacon for everyone roaming the desert at night. We designed a custom font and laser cut acrylic and MDF to make each letter. To mount everything together, we 3d printed connectors that would act as the walls of each letter.
As I mentioned earlier, there are a lot of options for controlling hundreds of LEDs. We’ve made many costumes and robots in the past, and each time we had to go through the same routine of wiring proto-boards, powering them, and figuring out how to attach LEDs.
After several prototypes, we came up with our own custom circuit boards. Using Fritzing to design the circuit, and PCB fabrication services like OSHPark, JLC, and PCBWays, we came up with several utility boards that allowed us to mount arduino style boards with lots of goodies like power management, screw terminals for LED strands, and a wireless chipsets to go wireless. For the brains we went with the tried and true M0 FRM69 feather by Adafruit.
“Did you just say 900MHz? Isn’t that like the old cordless phones? And why not 2.4GHz or 5GHz bands?”
We wanted to control our costumes and Bass Bar wirelessly and needed a simple way to do so. We chose the 900MHz band because of its lower power consumption and longer range than it’s GHz band counterparts. Higher frequencies are able to carry more data, but take much more power and have a shorter path. Here is a diagram illustrating how lower frequencies travel further.
Since we’re only transmitting a few bytes of data and not streaming audio, we wanted a longer range so we can jump off Bass Bar and party with the crowd!
The M0 coupled with the RFM69 chipsets were a breeze to work with and had many goodies out of the box. You can create multiple networks each with hundreds of nodes (255 networks consisting of 256 nodes each). You get 128-bit AES encryption and the ability to have peer to peer or mesh networks.
- Six unique custom circuits
- 12 Arduino M0 Feathers RFM69
- 1000s of SMD components
- Dozens of THT components
This is an area where our 30 years of combined experience really shined. We set out to build an immensely challenging project with so many interconnected components. Our previous lessons from years of leadership and management led us to balance what we built and what we bought.
We knew we had to build custom firmware and a custom network to connect all of our LEDs, but we decided to buy our DJ and VJ gear since we weren’t experts in that field. Plus, who doesn’t want to play with a Native Instruments Traktor or Maschine?
We also want to give a special thanks to the guys over at Resolume who generously helped us with some goodies, including their flagship software Arena — if you’ve been to any big show with big screens and graphics, then you’ve seen their platform in use.
Purchased tech aside, we did build a lot of software. We developed lots of custom arduino code to manage our network of nodes. We developed several LED effects which can be controlled wirelessly. We also developed “Light Painting” app to help us create light patterns live. ~10,000 lines of code later and Bass Bar was ready to come alive.
Come Find Us!
Making Bass Bar this year was our most challenging and fun project to date. Tim and I decided to build something that really pushed our limits and I hope sharing some of our story will encourage you to do the same.
After many cuts, burns, all nighters and patient neighbors who have grown tired of our sound tests blasting [insert the bassiest song you’ve ever heard — then imagine an even bassier song], we are ready to head out to the desert. If you find yourself at burning man come visit and say hello. And if you love engineering stuff like this feel free to reach out with questions or comments.