Why We Like Audio Watermarking

There are a number of ways to synchronize second screen content to the main program. BD-Live gives a two-way app sync for programs on Blu-ray Disc, made possible because the BD player and second screen device are on the same wi-fi network. Audio fingerprinting, made famous by Shazam, matches a segment of audio from the primary content against a vast database and serves second screen content accordingly. Video recognition is the next technology about to crest, operating on the same principles as audio fingerprinting, but theoretically would work even when the program is muted.  At CES in January 2012, Technicolor demonstrated a sync through the set-top box (the team behind this demonstration is now part of Magic Ruby). And zeebox has its own brand of sync, a “listening” algorithm that detects audio and closed caption cues and serves related content according to what it hears.

Each of these technologies has its strengths, holding advantages over other methods with certain types of content or playout environments. This is the main reason Magic Ruby is sync-agnostic: in a perfect world, the synced experience would be made available no matter the content, device or viewing environment.

That said, there is one sync technology that we find provides an excellent experience under the majority of circumstances: audio watermarking. Developed as a content security tool by Technicolor, the audio watermark transmits a “beacon” that keeps the app in sync with the main program.

The advantages to watermarking are many:

1. It works in all types of viewing environments, even with heavy ambient noise. I am a dedicated user of Shazam, but have you ever tried using it at a coffee shop when the steamer and grinder are on? At certain levels of ambient noise, the detector has a difficult time matching what it hears to the database, interfering with a consistent sync. This doesn’t happen as frequently with watermarking.

1. It works with DVR content, even weeks later. Fingerprinting a program requires that every second of audio be encoded and indexed in the database (otherwise, the app won’t sync for people who tune in late). This requires a significant amount of active storage and content management, so much so that most app developers flush fingerprinted material from their database within a week of first airing.  If you missed the episode from 2 weeks ago, you’re out of luck. By contrast, watermarking the same program leaves a much smaller footprint, since beacons are inserted only periodically throughout the content.  Thus, content can continue to sync as long as the customer desires.

1. It works with live events. Technology has evolved to the point that watermarks can be inserted during a broadcast, which is particularly apt for live sports and news. When tied to a dynamic content management system, you have the ability to serve relevant content to the second screen within moments – just like the statistics or news crawl that accompany a live match or newscast on the first screen. And the experience to the DVR viewer is completely seamless.

No technology is perfect for every case, but the flexibility and relatively lower demands on back-end architecture make watermarking an excellent overall solution.