17 years ago(!), I left a budding career in radio journalism out of disgust with the trajectory the industry was taking. The break-point came when the National Association of Broadcasters and National Public Radio teamed up in Congress to conduct a disinformation campaign designed to eviscerate the FCC’s then-newly proposed LPFM radio service.

However, A few months before I actually quit my job, I acquired all the components necessary to start an unlicensed microbroadcast station. “System P” was a 40-watt frequency-agile FM rig that used a portable military surplus antenna mast to conduct tactical broadcasts from a wide variety of locations. You could often hear the station in Madison, Wisconsin, primarily on evenings and weekends; but since the station was mobile much fun was had taking it to peoples’ homes and public events around the country to give the public a more substantive appreciation of the ease by which it could make “the public airwaves” very real.

Another key element of System P was to provide a last-mile node for what was then quite an experiemental webcast-activism scene (today commonly known as “livestreaming”). These often manifested in Independent Media Centers during times of protest, most notably against corporate global trade deals. Activists would converge on a city to fill the streets in order to disrupt the negotiation of these agreements, and the media coverage would invariably skew toward painting the activists as violent thugs and police/other security forces as the guardians of order. But when activists gained the ability to counteract this narrative – oftentimes by live reports from the streets directly – the discursive dynamic around these events changed.

An ad-hoc network of microbroadcast stations formed what was then known as the “Emergency Broadasters Bloc” to relay livestreams from these events to communities around the globe, for listeners who didn’t have Internet connectivity of their own. At System P, we would often supplement the livestream coverage with stories that provided more context to these events, and would host local activists (either at the protests themselves or engaged in solidarity actions) to help bring home an issue that, at the time, seemed overly abstract.

The station went with me when I moved from Madison to Champaign, Illinois to continue graduate school in 2004. But in 2005, after the Gulf Coast’s ravaging by Hurricane Katrina, I gave away my rig to friends who headed south to help with relief efforts. That electromagnetic incarnation actually helped forge bonds between disparate groups working the relief effort, and later moved on to service the New Orleans metropolitan area directly.

To my knowledge, the FCC was none the wiser about any of this – I never received any sort of contact from the federales over this five-year period.

Today, I work on a campus located at the geographical and electromagnetic epicenter of a renaissance in unlicensed broadcasting, the likes of which the United States has never seen. Last year, New York’s broadcast industry estimated that more than 100 pirate stations were on the air in the metropolitan area, outnumbering licensed stations in the market. And in the Flatbush neighborhood, to which Brooklyn College is adjacent, the vast majority of these stations are run by and serve a variety of Caribbean immigrant communities – most notably Haitians.

Radio has a long history of cultural primacy in Haiti, and Brooklyn is home to more than 60,000 foreign-born Haitians. It also boasts the largest concentration of Haitian-owned businesses in the country. Thus it is no real surprise that, since licensed radio stations never bothered to cater to this community and/or priced access to their airwaves far beyond what the Haitian disapora could afford, they took to the dial on their own, in contravention of the law. One Haitian pirate had a storied run of nearly 25 years here until they disbanded (for non-FCC reasons) late last year. Many broadcast only in Kreyol and/or French.

Brooklyn College is also home to a new Haitian Studies Institute, the first of its kind in America, with designs to become the premier research-destination for the study of Haiti and the Haitian disapora. There’s also the Brooklyn College Listening Project, an interdisciplinary endeavor now in its second year, which seeks to cultivate and archive an oral history of the borough’s vibrant cultural tapestry.

All of these elements are coalescing around an effort I’m devising to capture and preserve some of the radio created by the Haitian diasporic community here. The premise is simple and straightforward: design and implement a system to monitor and record “audio snapshots” of selected unlicensed broadcast stations for archival and further analysis. By its very nature, radio is transitory and ephemeral – you hear it and then it’s gone. Preserving some of these broadcasts would give scholars of the Haitian diaspora (and unlicensed broadcasting) a new window on which to explore the culture of these communities.

All we need to make this happen are “listening posts,” or monitoring-points where we can receive and record the snapshots of stations in action. Most pirates in the New York era operate on evenings and weekends, with Friday-Sunday being a “prime-time” of sorts. So we’re thinking of starting small, capturing audio during this prime-time, on a handful of stations selected by the Director of the Haitian Studies Institute for their programmatic value.

The first phase of this project is fairly linear and manual: acquire high-quality FM receivers, set them up in an area on campus to maximize reception of the selected stations we desire to sample, and auto-record a snapshot. Ironically, the best receivers for this task are no longer made: the quality of AM/FM reception in consumer electronics has actually declined in recent years, as manufacturers de-emphasize that functionality in order to make room for (and spend more R&D on) digital audio technologies such as streaming. There’s still a healthy resale market for them, however, though they aren’t cheap ($100+ per unit).

Each receiver would feed a bare-bones laptop solely dedicated to running a program designed to record a station during a particular moment in time, preprogrammed to do so. The open-source audio editor Audacity, for example, has an auto-record and auto-save function which would make this process pretty much automatic. Given that unlicensed FM broadcasts are not typically overly concerned with fidelity-issues, taking audio snapshots in something like a 64kbps MP3 file format would be adequate for the task; one 72-hour recording would generate a file about two gigabytes in size, but a single terabyte drive provides enough storage to archive a minimum of three stations for several years. Once the snapshots are saved, we’d then add the necessary metadata to the file and upload it to a dedicated server for archival and further analysis.

This first phase would simply be a proof-of-concept, making sure that we develop an efficient workflow for planning, taking, and saving audio snapshots. But it’s the second phase that will be most fun, as it inverts the notion of the “Emergency Broadcasters’ Bloc.” Instead of FM being used as a last-mile link for programming that originated online, what if we could use the Internet to extand the range of these stations, and our ability to listen to them?

The key to the monitoring and sampling-system is the Raspberry Pi. This credit card-sized chipset is essentially a tiny computer that can run a variety of flavors of the open-source Linux operating system. Raspberry Pis have garnered a large following in the open-source, maker, and educational communities since they were first introduced in 2012, as they’re cheap and eminently customizable to function as just about any sort of computing device you might want.

Some Pi enthusiasts have already designed complementary chipsets to turn them into software-defined radio receivers with the capability to scan a variety of bands, including FM. The software also exists to configure Pis into streaming relay-servers using the popular open-source Icecast protocol (the same platform we used back in the olde-tyme Indymedia days). In fact, there’s enough computing power and storage-space onboard a Pi to function as both an FM receiver and streaming relay. Perhaps most ironically, constructing and programming a Pi to work like this would not cost any more (and may even be cheaper) than setting up the first phase of this project, with its standalone receivers and laptops.

All each Raspberry Pi would need is a reliable source of power and Internet connectivity to act as a “listening node” to monitor any given Haitian station. At the appropriate scheduled time-windows, the node would activate its FM receiver, tuned to a single target-frequency, and then relay that as an MP3 audio stream to a central recording-server; no human intervention required save to add the necessary metadata to each recorded file. Because each node is compact and the streams would also be relatively low-bandwidth to a single client, they could conceivably be placed in any location where reception of the target-station is most optimal. And since each node’s design is relatively straightforward, once the first few go through beta-testing, building subsequent nodes would primarily involve piece-work and minimal tweaking.

Imagine a distributed network of listening-nodes such as this, scattered throughout the borough of Brooklyn, with the capacity to sample the entire range of the unlicensed broadcast ecosystem. Since all the necessary technology is off-the-shelf and open-source, the nodes would be available to anyone that might wish to similarly sample their local FM electromagnetic ecosystem. If the capacity existed to make access to these nodes public (not likely in the case of licensed radio stations), with the proper eye-candy front-end it would amplify a project like the Radio Garden by adding an entirely new dimension to the radio-listening experience – making the shadow-system of unlicensed stations that presently exist available to many more listeners than they could ever cover with an FM stick alone.

In essence, this project is a fine example of using old-school technology in a new-school way. But I’d love some reader-input on how you might go about setting up such listening posts. What do you think would be the most advantageous mix of technologies for both phases of this project? Are some radio-reception chipsets for the Raspberry Pi better than others for this task, and what programs would you install/configure on each node to make them maximally efficient and resilient?

There’s also a double-edged sword to this project – any network of listening-nodes established to collect audio snapshots of unlicensed stations could also be conceivably used as surveillance-tools, especially in an era where national-security and other “law enforcement” services in the United States may run rampant with few checks on their power. There’s even a move afoot in Congress to draft legislation that would criminalize the notion of “aiding and abetting” unlicensed broadcasters. Is there a risk in this environment that such studies of a media ecosystem could, in and of itself, be considered out-of-bounds?

Feel free to drop a line in the comments, or sound off directly. I’ll provide updates to this project as it progresses.