Monday, October 26, 2009

Text to Speech Functions as a Podcasting Tool

Text to speech software has been evolving for decades, though there has been substantial progress in recent years. For those of you who have purchased a Mac computer in recent years, the compute comes with a text to speech system that can be used to read text.

I've run a podcast for several years at my airline safety site, and last week I decided to run a test to see how well the audience would accept a computer voice instead of a human narrator. I combined that with the online form I talked about in the last post to do two things: try out the text to speech software as a narrator and evaluate the audience's reaction to it.

Early results show that the podcast, which was just under five minutes long and was about two airline safety incidents from late October 2009, was generally well received. Most of those who responded listened to the entire podcast, and also felt that they could stand to hear the voice for short episodes. Surprisingly, most said they would listen to shows of any length.

One of the options for this text to speech system is to supplement regular interviews. There are dozens of comments and emails that come in during the typical month, and some would be appropriate for some podcasts. Also, those who wanted to submit guest commentaries but who did not have recording capability could write their response, and the software could read it.

For the aviation safety podcasts, the ultimate goal of using text to speech software isn't to avoid recording regular voice conversations, but to provide a way for the portions of the audience without podcasting equipment to contribute to the ongoing conversation.

Listen to the Podcast Episode
Northwest Airlines A320 Overflight Event and the Delta 767 Landing Incident

No comments: