Jeff Jarvis at Buzzmachine writes about the possibilities of using closed-caption data associated with a video feed as a source of searchable information. In particular, the time-synchronized nature of the caption information would allow linking to specific sections of a media file, from a text search.
Chicago Captioning Corp. added closed captioning to the video.
They did that in an effort to serve the 10 percent of Americans who are hard of hearing. And that’s great.
But I see another important use that is of value to 100 percent of Internet users:
By attaching a script to the video, we get metadata associated with it. That makes the video searchable via Google et al. That means that the content of the video can be analyzed. That means we can link to specific content.
This sounds great, but closed captions are only available for video, right?
Well, no. MP3 audio files have had “lyrics” support for a long time, and a surprising number of players will play “karaoke” audio files. And “karaoke” is nothing but time-synchronized text associated with an audio file, exactly like closed-caption data in video.
While I’ll agree that supporting closed-caption metadata should be a priority for video-blogging software, it’s currently a reality that there are far fewer videoblogs being produced than audio-only podcasts. Supporting “lyrics” (a.k.a “show notes") and “karaoke” (a.k.a “transcript") metdata in podcast creation, distribution and player tools would be a much more significant step.
It seems to me that there is even an “organic” way to achieve this. If podcast creation tools routinely supported display of “teleprompt"-style notes while recording, it would encourage people to enter their notes that way rather than fumbling with paper or a wordprocessor. Once the text is in the creation application, and scrolling in time with the input, it’s a snap to output synchronized and unsynchronized metadata to the recorded file. No tedious manual synchronization or after-the-fact transcription needed.
It’s my most popular rant at the moment. Audio creation, distribution and player applications are really missing out on the richness of opportunities in metadata. So much that people using these systems don’t even realize that things could be so much better.
To quote jeff Jarvis again:
Metadata, man, metadata.
Read more at BuzzMachine… by Jeff Jarvis