Fifteen years ago, some friends and I started thinking about how listening to music was going to change. That was 1999, before streaming services or even the iTunes Store, so music collections were made up of CDs and files downloaded from early peer-to-peer networks.
It was a time of tremendous promise. Music, which had long been bound by the physical discs on which it was sold (and the broadcast media which promoted it), was going to become massively accessible. It was an intoxicating thought that all the music in the world might actually be something we could see and hear in our lifetimes.
Less exciting was what seemed to be an inexorable trend: as music became more convenient to acquire and hear, its presentation became less and less compelling. In the 70s, a gatefold sleeve for a vinyl LP had huge full-color panels of artwork and photographs, plus inner sleeve panels for lyrics, credits, and other information. In the 80s and 90s, CD jewel cases cut the outer surface area of the packaging by 80%, and you needed reading glasses to make sense of the liner notes. By the time digital media players surfaced, music was reduced to a list of file names.
Somewhere along the line, it was accepted as law that the world of music — resplendent in its diversity — could adequately be represented as a spreadsheet of artists, albums, and tracks. We, as certified music nerds, weren’t prepared to live with that.
Because of that trend, we weren’t contemplating the practical problems of how music would be distributed, stored, or streamed (or who would sell it or how much it would cost) but what the experience of a limitless music collection could be like. What was exciting to us was not the idea of having all the music in the world, or even having access to it, but how we might interact with it. What could we learn? What new music could we discover? How could we understand and appreciate music even more deeply?
The early years
We kept our day jobs and experimented, building software that we hoped would lead us to our holy grail: the rich experience of browsing LPs combined with the emerging digital promise of access to all the music in the world. We felt strongly that the experience should include a graphically rich touch user interface, be as fast as a modern video game, contain all the metadata we could get our hands on, and support a huge library of CD-quality music.
After two years, we succeeded. We created an immersive music browsing experience that didn’t exist anywhere in the world… think High Fidelity meets Minority Report. Unfortunately, it needed a small truckload of hardware to run on, because the consumer PCs of the day just weren’t powerful enough.
By 2003 we had our first test system (affectionately known as “The Bitch”), a 19″ touchscreen kiosk with a gaming GPU, a ton of RAM, and a two terabyte array, built by hand from $20,000 of parts. It required more than its share of care and feeding, but it was a start. We loaded our music collections onto it and dragged it around New York City to parties, where people would invariably huddle around it and get into heated discussions about music.
Eventually the component costs came down a bit, and we installed smaller units at two of our favorite bars. One night, I watched a couple get into an argument because she wanted to go home and he wanted to stay and mess around with the “music machine.” I called my partner Danny Dulai and we agreed that we knew what we wanted to be when we grew up.
The triumph of reckless optimism
We’re software guys, and we had built The Bitch out of necessity. CPUs, RAM, and disks were getting cheaper, and there were signs that touch computing would be going mainstream, but at that point we couldn’t just give our friends an app to download. Instead of doing the reasonable thing and waiting for consumer hardware to catch up, we did something insane: our ragtag bunch of music and software nerds decided to start a company and manufacture hardware. Lots and lots of hardware.
I went to Taiwan, met contract manufacturers, learned SketchUp, and began the long process of figuring out what it takes to make physical objects out of nothing. We produced many disastrous industrial designs and botched everything we tried at least once. When our first prototype arrived in New York, Danny sent me an email with a simple subject:
“We are going to fail.”
It took us nearly two more years of trial and error and more than one trip back to the proverbial drawing board. We tirelessly solved problem after problem and in the process, we paid more and more attention to increasingly subtle details. Our unwillingness to compromise caused us to begin to understand the art of moving from “concept” to “product.”
In 2006 we launched Sooloos, and despite its hefty price tag of $12,900, it was received by customers and reviewers with open arms. The product was a distributed network system of control, storage, and player devices, which could be configured to provide up to 32 zones of music and multiple terabytes of redundant storage. Under the hood, we had designed protocols for device discovery, configuration, and streaming, written a game engine for the touch GUI, created a lightning-fast database, built a cloud service to aggregate metadata, and completely rethought every aspect of audio encoding and decoding. And we got the boxes right, too.
With hardware design and manufacturing apparently under our belt, we spent the next two years learning about inventory management, logistics, distribution channels, margin structures, and service and support. We attended trade shows we had never heard of, did interviews with journalists from magazines we didn’t read, and referred to ourselves as “teenagers talking about sex,” which fairly described our fumblings in this mysterious new world of hi-fi.
Eight years into our adventure, we had learned a ton and had more fun than we could have imagined, but we realized that 90% of our effort was going into designing metal boxes, getting them built, shipping them from A to B, writing invoices, and providing home networking support. We had been so busy that we scarcely noticed ourselves drifting away from the goal we set at the outset: to connect people with music.
In 2008 we partnered with Meridian Audio, because we understood that to further our plans for world domination, we needed real expertise in manufacturing, distribution, and above all, audio. Meridian’s pioneering work in digital audio and innovative approach to product design made the companies a perfect fit, and Sooloos was acquired by Meridian at the end of that year.
Over the next few years, we made some significant improvements to the Sooloos products, notably integrating streaming services so all of a user’s music — from CDs, downloads, and streams — could be viewed side by side with all the rich metadata we had become known for. With Meridian’s help, we also took the leap into synchronized multi-zone streaming, high-quality sample rate conversion, and DSP. Sooloos had become the gold standard for high end digital music systems.
Getting to scale
As the cheap PC, NAS, smartphone, and tablet became ubiquitous, we took our distributed embedded system and began to repackage it as a standalone piece of software. We didn’t yet know how it would be marketed, but there was a long road from where we were to where we wanted to go, so we started a skunkworks that would form the foundation for what we do today.
In parallel, Meridian had been introduced to HP, which was reworking its strategy for music and media experiences at the time. HP asked us to develop a branded app based on the Sooloos work we had been doing, which resulted in the launch of HP Connected Music in 2012. All at once, we went from serving thousands of customers a year to serving millions every day, so we earned yet another badge: competency in the cloud infrastructure necessary to serve this much larger new audience.
Onward and upward
Today the world is changing yet again. Music services are expanding their offerings and there is more affordable high-quality audio hardware implementing streaming than ever before. We believe that people still love music and want to be connected to more than what’s “trending,” so that’s what we’re going to do.
In Feburary 2015, our original Sooloos team reached an agreement to spin out of Meridian to form Roon Labs. We will be launching our new product in May, and next week I’ll tell you a lot more about it.