[Interviewer] Henrik Moltke, you're going to tell us about your Hyper Audio project.
[Interviewer] What is it, if we start at the most basic level?
[Henrik] Well, it's really just combining hypertext, which we know from the internet with audio, or plainly sound. I used to work here at Harddisken as a reporter and have worked with radio and sound for many years. One of the things that was a problem was that it's very hard to make sound and the internet work closely together. When we put radio out on the web we start from the metaphor of taking a physical radio and then placing it 'on the web' and then you can press play or stop. But that's not how the web works. So the idea behind this project was to find ways to put the web more into radio, and vice versa. There are some very low hanging fruits when you do that. It's hard to search audio it's hard to navigate within sound on the web Audio is isolated and insular You cannot easily navigate between different kinds of players on websites The audio is 'flat' when on a webpage, for example in a Flash player. So is an attempt to to insert the net into sound.
[Interviewer] um -
[Henrik] I have to rephrase that: It's an attempt at getting radio more into the web and the web into radio. Allow me to interrupt you preemptively:
[Henrik] Recent developments with the native web language, HTML In the latest version of HTML, HTML 5, you to can put audio on a web page without using an external player, and do thing where other page elements, and the rest of the web can merge with the audio. This is what's called the audio tag. There's also a video tag. At Mozilla we have been doing some experiments to make this reach a higher level.
[Interviewer] Before we get more into technical details, we have cheated a bit. Like in a proper radio tv kitchen you have prepared something based on existing Harddisken content Can you, even though it's very meta, something on computer which we talk about in the radio and is about our show here - could you demonstrate it?
[Henrik] Sure! In December you interviewed my colleague Tristan Nitot, president of Mozilla Europe. I took this segment and re-did it in two languages. I've started playing it here, and as you can hear this is Føhns. If I scroll down and click, and here you are you recognise the voice?
[Interviewer] Yes, this is me speaking the segment from March.
[Henrik] And if I jump to the english version - there's a button that toggles language.
[Interviewer] If you stop it for a moment.What you have done here is switched to an English version of my voice-over?
[Henrik] Correct. This page has both subtitles and multiple languages. But the real killer feature is the script which is based on what you originally shared with me We have made each word a link. So if you switch to the danish version and hit play I see that somewhere in the script here it says 'nogenlunde stabil' and if I click there it goes to that place.
[Interviewer] So what you see on this page is the entire text from my manuscript and you can click any random word, and the audio file start playing from there.
[Interviewer] That means you're not forced to listen through a 20 minute segment to find the spot when, in this case Tristan Nitot, is talking about the subject, you were looking for.
[Henrik] And when the audio is hyperlinked the data available for search machines becomes more granular. Which makes it a lot easier to find something you heard sometime in the past, on the radio.