Preparing a video speech
Once in a while I get the chance to nerd out about a particular aspect of my work. And in the case I want to describe today, there was even a crossover with my hobby, podcasting!
I recently interpreted at a high-level event with both broadcast and web coverage. One speaker, however, could not attend in person, and was given the opportunity to record their speech in advance as a video. Thanks to our head of team, we even received the video beforehand! A video, however, presents different challenges from a written manuscript, such as linearity and dependence on sound.
This is where my podcasting experience came in handy. Having the video in advance was great, but I wanted an audio version to be able to listen while out and about, and I also wanted a manuscript of the speech. Auphonic to the rescue!
Auphonic is a web-based service from Austria that, broadly speaking, improves audio and also converts many video and audio formats. So not only was I able to extract the audio from my video speech and download it as an MP3 file, Auphonic also connects to Google's speech recognition engine. That means, quite simply, that you can have speech recordings transcribed automatically. The results aren't perfect (and will depend on the language), but they're definitely usable to find one's way around a transcript.
But wait - there's more! Auphonic does not only export the Google-powered transcription as a neat HTML file with time codes, but also a VTT file (the standard format for subtitles). So with an app like VLC, which is available for Mac, iOS, Windows and Android, you can add those subtitles to the original video!
Yes, this is nerdy, but it shows that technology can be very helpful in supporting our work.