WAMU Teams With ENCO For Hearing-Impaired Connectivity

0

WASHINGTON, D.C. — When it comes to texting and hand-held communications, there’s perhaps no group of National Capital Region residents more experience and savvy with such devices as the Gallaudet University student body and faculty.


This Northeast Washington institution has provided higher education opportunities to the deaf and hard-of-hearing for generations. While TV has been, along with print media, major services consumed by the hearing impaired, radio has been left out.

American University‘s NPR Member station, across town in Northwest Washington, has changed that — thanks to ENCO.


BE SURE TO FOLLOW RBR+TVBR ON FACEBOOK!


To bring WAMU-FM‘s programming to all Washingtonians, the station deployed an enCaption4 captioning system from ENCO.

This generates live, automated transcripts of its on-air audio via the WAMU.org website — not so dissimilar to the audio transcripts of FCC open meetings found on the agency’s website.

As of today, live on-air captions remain in beta test. “We’re working to add live real-time captions to our radio broadcast,” WAMU says. “While we continue to refine this early release effort, please pardon our errors and inaccuracies.”

Interestingly, the caption feed is several seconds ahead of the audio streaming feed; this suggests the captions are synched with the over-the-air broadcast.

The inspiration for WAMU’s captioning project came when a deaf political candidate seeking office in Washington requested to appear on WAMU’s Kojo Nnamdi Show and wanted the broadcast to be accessible to all.

“At that moment, it stood out to us that we hadn’t been serving that important part of the population before and that there was demand for bringing our content to this audience,” said Rob Bertrand, WAMU’s Senior Director of Technology.

For that particular broadcast, the station used a service provider with live human transcription and live sign language interpretation for that particular broadcast. From there, Bertrand knew that a different approach would be needed to serve the hearing impaired community on an ongoing basis. “The timing of the content can be unpredictable, which means we can’t schedule a transcriptionist or sign language interpreter,” Bertrand said. “We realized there was now technology that would allow us to automate the process. Automated closed captioning has existed for television for some time. What if we could take the equipment used to generate captions for TV, and instead use it for radio to bring captions to our website?”

The enCaption4 was selected by Bertrand after learning of lots of positive feedback from broadcasters using it for television. “We did a demo of enCaption4, and our web development team came up with a way to present the captions it creates on our website,” he said.

How does it work?

Audio from WAMU’s Telos Alliance Axia audio-over-IP backbone is converted by an Axia xNode to an AES/EBU signal that serves as the input to enCaption4. Then, enCaption4 ingests the same on-air signal path being routed to WAMU’s transmitter and online streaming encoders, enabling live, 24/7 captioning of all of WAMU’s on-air content. The captions created by enCaption4 are then fed to the station’s website, where they are displayed in a dedicated transcription page.

As a pilot project until now, WAMU developed a small but loyal following, with roughly 150 people routinely watching the station’s captions for around 30 minutes per day. WAMU expects this to grow once the service starts getting promoted.

Furthermore, WAMU plans to upgrade its content management system to allow integration of its internal data sources with enCaption4, thus providing a dictionary that will further increase the accuracy of WAMU’s captions with the spellings of challenging local names and events.

The station also ultimately intends to integrate the captions directly into its core streaming player.

— Reporting by Brian Galante. Editing by Adam Jacobson.


Adam Jacobson is a 1994 graduate of the American University School of Communication.