The Digital Public Library of America partners with Pop Up Archive

Searchable sound for libraries across the United States


Libraries across the United States house tens of millions of audio and video recordings, a rich and vibrant body of cultural history and content for the public, scholars, and researchers — but the recordings are virtually impossible to search.

The Digital Public Library of America is partnering with Pop Up Archive to offer discounted services to the DPLA network. This represents 1,600 libraries, archives, museums, and historical societies across the US. Through new service offerings available exclusively to the DPLA’s partner organizations, Pop Up Archive will automatically transcribe, timestamp, and generate keywords for the audio collections that take advantage of our partner discount.

Continue reading

Opening up audio archives through human-­computer collaboration

Pop Up partners with The New York Public Library and The Moth to build a new model for accessible audio

Screen Shot 2015-11-10 at 1.10.09 PM

We’re excited to announce our part in an innovative audio search project that combines two of the things we love most: libraries and storytelling. Thanks to the generous support of The Knight Foundation, The New York Public Library was awarded a grant in partnership with The Moth and Pop Up Archive to prototype an automatic transcript generation and correction model.

Continue reading

Can personal taste power podcast discovery?

How can human curation help audio discovery at scale?


We spend a lot of time thinking about searchable audio, and we know that nothing is more daunting than a blank search bar. For every person on a hyper-specific quest for a certain audio story, there are many more who can’t articulate what they’re looking for — or aren’t even looking at all. When it comes to audio stories, we’re picky about the voices we’ll let yammer through our earbuds.

We don’t want recommendations that are robotic or anonymous. The most exciting recommendations have a “right place, right time” serendipity. What’s the best way to model that discovery experience for thousands — even millions — of users?

Continue reading