How about an app that does both?
I'm pcturing an app that you can switch back and forth freely between
audio and text. Say you're reading a book, you're in the middle of the
chapter, and your eyes are tired, or you have go mow the lawn, or you
have to drive somewhere, but you want to continue with the story. In
this app, all you would have to do is click on a paragraph, and the
audio version of the book would pick up right there iohtout skipping a
beat. when you're ready to start readig again, just click the screen
again, and you're back in the text.
The tricky bit of this app us that you'd have to have the sound files
marked with the start of each passage. I picture combining Librivox's
service where volunteers read books outloud, and stanza, where readers
can DL ebooks for free.
The trick is how to implement the design. So let's assume that you
have a public domain book in both audio on librivox and text in
another public domain database, like how Stanza stores its books.
theoretically, the books have already been broken up into chapters
electronically, in both formats. On Librivox, books are records by
chapter, where each chapter is a separate mp3 file (or other file
format). Therefore, the text and audio formats of a book could be
linked in the way described above, chapter by chapter. This idea could
be implemented right now on that basic level.
However, this is not the level of synchronicity I envision. Rather, I
see text and audio being linked on a sentence for sentence basis. This
would require a greater amount of work from volunteers, who would have
to somehow manually link each sentence of text to the spoken sentence.
I see this being accomplished in one of three ways. Each is not
without its difficulties.
One, somehow incorporate a voice-to-text service such as the ne used
in Google Voice transcriptions. The difference would be, the service
would have the text already, and would merely have to match up the
sounds it was hearing with a specific passage of the text.
Two, when volunteers record their chapters for Librivox, find some way
of electronically marking each sentence in the sound file. The app
would be able to find those markers and then match the correct passage
with the correct portion of the audio file. This could be made easier
by utilizing the audio recording capabilities of the iPhone. Readers
could touch text on the screen with a finger as they read, giving the
app markers to link the audio and text.
Three, find some way for other volunteers (after the recording has
already been made) to link text with audio. I envision a program, or
different mode in the app I'm describing here, where readers follow
along with the audio and make links in the text to the audio's current
position. This may be my favorite option, because it could simply
build upon material that has already been created, without having to
re-record the audio. We could possibly have something that listed the
reader's "stats", like how many markers they did that day.
I wish I had the iPhone programmng skills to be able to program
something like this.
No comments:
Post a Comment