By Mark Boas
Sometimes when we take on projects we don’t really know what we’re letting ourselves in for. To fully know, we’d spec things out to the ninth degree and who wants to or has time for that? Well people like NASA do, but we don’t.
When we were given the opportunity to work on something called Perceptive Media all I saw was a colourful but amorphous form ahead and although it was explained to me (as well it could be) I still didn’t have a real clue of what it would end up being and crucially what was involved in making it a reality.
But hey it’s the BBC right? You don’t often get to work with the BBC – you know that BBC you spent so much time watching and listening to when you were growing up. The same BBC who make Doctor Who ? – That BBC!
Time-lords aside (or not actually as the case may be), being a bit of an audio-geek I’ve always been fascinated by the BBC’s audio output especially their Radiophonic Workshop especially Delia Derbyshire’s work and this was an audio gig coming from the R&D department of the same organisation. How could we turn it down?
So I jumped at the chance without really knowing what lay in store, which incidentally is very unfair on my colleague who ended up doing the lion’s share of the work.
The brief went something like this : We want to create a web-audio based demo that will adjust its content to the listener, based on information we can ascertain about them. Oh and we want it to sound as natural as possible so that the listener may not suspect that the content is being tailored. I soon found out that this involved generating audio on-the-fly and applying convolution reverb and other effects so that audio sounded natural in various environments. What this all meant is that we needed to use an advanced audio API. Now I’ve had fun investigating advanced audio APIs before, they differ from the standard audio APIs by being designed to allow you not only to play audio files, but to generate and alter audio.
This is exciting because we are finally getting around to doing stuff with audio that we have been doing with text for years. It also crosses over with something I’m looking at called Hyperaudio.
So with this cross-over in mind, I felt I could manage to find the time to experiment and create some proof of concept demos and this lead me to try out the following libraries:
The Speak.js demo pretty much demonstrates what it can do. You put text in and speech comes out. The library itself is ported from eSpeak using something called Emscripten and actually allows you to generate audio pretty much in real-time by constructing a data URI in WAV format.
Unfortunately time and budget restrictions meant that we couldn’t create the version we wanted to make – a version that would work similarly in all browsers that provided an advanced audio API. As it seems that the future W3C standard for advanced audio will be based on the Web Audio API we decided to concentrate on that.
The data we use to personalise the broadcast comes from a number of sources. The main differentiator is the listener’s location which we use via the geo-location API to determine, local weather, radio streams and landmarks which are then subtly inserted into the audio stream. The only restriction is that you must be in the UK to really encounter the differences – this is partly due to the fact that we use some BBC resources that are only available for the UK but also so that we can keep the data manageable for what, after all, is just a demo.
The second source of information about the listener came from a slightly more sinister place. For fun I’d been working with a good friend of mine Matteo Spinelli on a project called Underpants and while looking at the issue of browser traceability across websites we figured out how to determine which social networks a browser is logged into. Cross-over struck once more and we used this technique to personalise the part where our outdoor-challenged hero is urged to log out of her favourite social network and leave the apartment.
Advanced Web Audio
So what changes between the version of the broadcast that uses the Web Audio API and the fallback version that doesn’t? Well there are a number of factors, some of them quite subtle. Speak.js outputs a robot style voice which is fine for our use as the electronic voice of the lift. But we wanted to make sure it would fit in properly to the various environments in which it was set. To do this we created something called a convolution-reverb. In short, a this reverb allows us to apply the right sort of audio ambiance to a sound. So if a sound is coming from a lift we apply a lift type echo. We also apply the same ‘echo’ to the streamed radio broadcast that is played at a certain point in the broadcast.
The fact that we are using an advanced audio API also enables us to add various other effects to other pieces of audio. However we soon found that we needed to be sensible with our audio design, since convolutions with the Web Audio API does take a up a fair amount of CPU. During a development error, a unique convolution was used for each sound, this was found to start failing at around the 14th.
We also made use of audio filters, for example the radio podcast uses a high-pass filter applied to it which makes it sound ‘tinny’, another example is when Harriet opens the her apartment door at the start we apply both filters and faders. The Web Audio API uses a node based approach which means that you can feed the results of one effect into the next so we can apply filters, faders and convolutions to any audio source. To achieve all of this we made heavy use of Web Audio API’s AudioParam which allows nearly any attribute to be changed using handy linear transform effects – we used this to fade in and out, or cross-fade between filtered and unfiltered outputs.
So the Web Audio API version applies filters, faders and convolutions to the audio whereas the standard HTML5 audio versions do not. That’s not to say that given enough time we couldn’t have achieved the same effect using the Audio Data API included in Firefox. But since the new audio standard is slated to be largely based on the Web Audio API it was decided for the purposes of this demo to concentrate our energy in this area.
Once we’d got the core of the functionality working, we set to work on creating a control panel to allow us to tweak every single one of the volumes, filters, faders and convolutions.
We wanted to be able to demonstrate to editors of audio how we could tweak pace, reverb and sound-effects in real-time and although requiring a complete code re-factor we hope that the fact that the whole thing is pretty-much customisable makes this a powerful demo and will be useful to others as well as ourselves who are dabbling in this area.
So this is a good time to mention that all the source-code is open source and that you can grab it from the following GitHub location https://github.com/happyworm/PerceptiveMedia.
I think this was definitely an interesting and worthwhile experiment. However as its aim was to be subtle, it purposely does not make immediately clear the potential of the technology. Technically what we are able to achieve turned out to be a kind of audio framework to allow the ability to create and tweak audio as it’s being played. This is useful to producers of audio to see how effects and timings can alter the experience and is especially useful for applications such as games where perhaps you want to give your sound-effects context. I also feel that these techniques could be used for applications such as dynamic story-telling. My daughter — all too often — asks me to tell her a story featuring robots, dinosaurs and goblins and all too often I fall back on the same old principles and formulas of children’s story-telling, the rule of three and so on. Post happy-ever-after she often wants me to add to the story, getting me to fill-in or clarify some of the details. It’s not a huge stretch to imagine that we could create dynamic storytelling applications for kids. A pinch of AI here, some personalisation there and a heavy dose of randomness might just be enough to keep them happy for a bit.
So there we have it, one small step closer to the old Star Trek computer (Did Doctor Who’s have a voice interface? I forget). We’ve already seen the application of voice input with software such as Siri. It shouldn’t be too much longer until audio interfaces start to become common-place, with so much current emphasis on the visual I think this could be quite refreshing.
Question. What’s harder to debug than an intermittent bug? Answer. An intermittent bug that only manifests itself when you deploy to the server. Crazy I know and totally unexpected to us and for this reason you may see issues when running on Firefox (but not Opera). Being supporters of both Mozilla and Firefox we were much dismayed by this bug and spent a significant amount of time trying to get to the bottom of it. Unfortunately due to its nature we were only able to put in a loose bug-report If anybody wants to help us solve this issue please feel free to take a look, even if it’s just to download the application from GitHub and verify that it works locally for you.
Thanks then to Ian Forrester and Tony Churnside for the opportunity to work with them and their team at BBC R&D, also of course Sarah Glenister for the excellent script and Angie Chan for the great artwork. Jussi Kalliokoski for helping us work with AudioLib.js But most of all I want to thank my colleague Mark Panaghiston for working tirelessly behind the scenes not only on the significantly challenging audio aspects of the project but also go above and beyond in integrating the visual aspects and even sourcing and setting up the hosting.
Writing for Perceptive Media
Illustrations for BBC R&D’s Perceptive Media Demo: Breaking Out
What is Perceptive Media?
BBC demonstrates revolutionary ‘perceptive media’
Perceptive Media Launch at Social Media Cafe Manchester
The BBC unveils its first ‘Perceptive Media’ experiment – and you can try it now
What is Perceptive Media?
This year has run contrary to previous years of my life in the sense that it seems to have lasted much longer than the year before. This may have something to do with a small addition to the family that arrived in March and the consequent ramping up of waking hours, but I prefer to think it’s because things are moving so quickly in that wonderful place we call the worldwide web, that it seems that too much has happened to compress into a year.
A few months later I was introduced to and became involved with something known as Hyperaudio and was given the opportunity to create a couple of proof-of-concept demos : Denmark Radio’s Hyperdisken Demo and the Radiolab/Soundcloud collaboration ( a shout-out to Henrik Moltke for doing much of the groundwork for these demos and my colleagues at Happyworm without which they just wouldn’t have been possible).
Further on through the arc of this epic year, I took part in the Mozilla Knight Journalism challenge where I was encouraged to research and blog some ideas on a tool I called the Hyperaudio Pad. Happily I was flown over to Berlin with other like minded MoJo peeps, for a week of intense discussion, collaboration and hacking and managed to get the first semblance of an actual product out of the door.
Which brings me in a round-about way to this year’s Mozilla Festival, where I will again be demoing and chatting about the coupling of text and media and the other things that Hyperaudio is (or could be) about. This all at the Science Fair on Friday, 4th of November and a Hyperaudio Workshop / Design Challenge on Saturday, 5th.
The Science Fair will be all about demos, chatting to people interested in the general Hyperaudio concept and making sure I don’t spill my drink over my laptop. Meanwhile in the workshop we hope to put the Hyperaudio Pad through it’s paces and actually make something with it, while gathering ideas and maybe hacking on some of them into the bargain. It should be a fun mix of hacks, hackers, designers, audio buffs and basically anybody curious enough to be involved. The session will be fairly free flow but the plan so far is to:
- Briefly chat about and demo the Hyperaudio concept, what it can do and with whom we can collaborate.
- Breakout into groups to come up with new ideas, applications and designs.
- Reconvene and talk about the group’s ideas.
- Breakout into groups to work on these ideas, whether it is creating a programme, UI mockups, story-boarding functionality or even developing small demos.
- Creating something we can show other people – hopefully a program created with the Hyperaudio Pad but also some new ideas.
Hopefully we can cram this all into 3 hours. Folk from the BBC, Universal Subtitles, Sourcefabric and other cool people are planning to be in attendance, which should make for some interesting crossovers and fantastic opportunities to consult and maybe even influence organisations in a position to use the stuff we’re making. Additionally the BBC are providing us with some fantastic media to work with.
All will be revealed over the weekend, book your place and get your eyes and ears ready to be part of the Hyperaudio experience.
Following the fun we had making the Hyperdisken demo, I was happy to be asked by Mozilla, in collaboration with Radiolab and SoundCloud to help create another demo to show off the possibilities of hyper audio. This time we had an excellent Radiolab program as audio material and we wanted to get a little more ‘involved’. What was required was an application that would consist of many of the features of the Hyperdisken demo but also integrate deeply with SoundCloud API, and on top of this something extra, something to catch the eye.
I was fortunate again to work with the ideas-forge known as Henrik Moltke, who collaborated early on with Paul Rouget to produce something he dubbed the ‘Word River’ – a CSS3 manipulated flowing river of words that dynamically picked up content from an HTML transcript. We were also keen to make a pure HTML5 based solution and Paul helped figure out the hooks into the SoundCloud API that would allow us to achieve that. We were also very lucky to be given a great design by the multi-talented Lee Martin, SoundCloud’s experimenter extraordinaire.
So with proof of concept and some visual bling firmly in hand I was tasked with making this baby fly. Luckily I had help. SoundCloud engineers were at hand to answer any questions and crucially we had great support and code contributions from the popcorn.js group. I also managed to talk jPlayer author and all round JS media guru and of course, colleague Mark Panaghiston into giving me a hand. So despite the tight deadlines we were pretty much set.
Henrik has already blogged about the ideas and functionality that make up the demo. I want to write a little about the technology used.
Although I found out in retrospect, not strictly essential, we once again used jPlayer as our audio base, we’re familiar with it and we can move fast using it. It also meant that we could take much of the functionality developed in previous demos and plug it right in. Again, the excellent Popcorn.js was the engine that drove all the time based display of text and images and dealt with the parsing of data. Steven Weerdenberg, active Popcornista from Seneca College, very kindly wrote a plugin that grabbed, parsed and presented comments (amongst other things) from the track we used hosted on SoundCloud. This is where the Popcorn framework comes into its own as a plugin oriented architecture, something we took advantage of when we converted both the transcript and the word river functionality into plug-ins.
So where did the data come from? Well again the transcript HTML doubled as the source for richer interaction when used by the word river plugin. I like this approach I have to say. It means you don’t have to be a programmer to come in and immediately understand the content and change it accordingly. I also like the fact that the transcript is a separate HTML file, it pleases the separatist in me and means that it works as a standalone resource. We also used the standard speaker notations as a type of meta-data, the word river plugin hiding these parts for the purpose of display but using them to colour code each speaker’s text.
This part of the transcript:
is used to create this ‘word river’:
and this interactive transcript :
Data-wise, everything else came via the SoundCloud API, this included their trademark wave-form, both ogg and mp3 audio sources and all of the comments. We also hijacked the comments to make a crude content management system. The idea being that any comments posted by the Radiolab account with references to images in them, were picked up and displayed as images in the main content area, and did not show up as comments on the timeline. If two images were present it meant they were square, one and it was ‘widescreen’ a blank image was used to remove images when they were no longer needed.
The last pieces of the puzzle and one we’ve still some polish to apply to, (if polishing puzzles makes any sense to you) was getting it all working on the majority of tablets and mobile devices. Since this demo didn’t use Flash this was actually a possibility and we got our web designer Silvia Benvenuti to come out of maternity leave and sort this out for us at the 9th hour, leaving me quite literally holding the baby.
This was a tough gig but all in all I’m happy with what we achieved, everyone seemed to really enjoy taking part in the process, and I certainly enjoyed bringing it all together. Hopefully it will inspire both program makers, designers and developers to come together and explore the limits of what hyper-audio can do. As Inspiral Carpets would say, moo!
Source code for this project and other demos can be found on github
Follow me on Twitter if you want to hear more about this sort of thing.
Recently I had the privilege of working on a very interesting project with a few folk from Mozilla – it’s the type of project I love to work on, as it involves web audio and its deep integration into the general web experience.
Web audio is no longer consigned to being the passive play and pause experience of yesteryear, it has the potential to be much more, it can be a driver of much richer interactions, something Henrik Moltke explores with something he dubs Hyper Audio. The remit of the project was to take various media elements of a radio interview broadcast by Danish Radio station DR; audio, subtitles, transcripts, footnotes etc and link these in an intuitive and useful manner.
To say this project was right up my street would be an understatement – this project was in my flat, raiding my fridge and drinking my beerz. I was already fascinated by the concept. I’d been playing about, creating audio related demos for a couple of years and in November last year I decided to attend the Mozilla Drumbeat festival and created a demo for the event. The demo was accepted to be exhibited at the science fair on the opening evening and garnered some interesting feedback both on and offline, what it effectively demonstrated was the synchronization and bi-directional control of text and audio.
When Henrik asked me to work on this project, I naturally jumped at the opportunity. Due to time differences, pressing deadlines and the luxury of having a nice quiet office, I stayed up late most nights for a week, happily hacking away and helped out and supported by various Mozillians and the popcorn.js community.
So that’s the back-story, here’s the demo.
Some things to try :
- Switch the audio from English to Danish – it should continue from the same point in Danish, subtitles and the transcript should also change appropriately.
- Try clicking on words in the transcript – the audio should start playing from the corresponding point.
- Highlight a passage of transcript text – this should add a tweetable excerpt to the ‘share’ box. The URL included should just play that part of the audio.
- Clicking the music note icons in the ‘media’ box should take you to the point of the audio where that resource was mentioned.
How did we achieve this? We used popcorn.js to display subtitles, footnotes and other time-related resources. In fact a lot of this was already in place when I picked up the project. I then integrated jPlayer for the audio playback and deeper interaction. Popcorn allows us to associate timings with actions and have these actions triggered by media when they hit said timings. So pretty much perfect for our needs. jPlayer provided a solid abstraction above the native audio API, it allowed me to easily synchronize and switch audio tracks and jump to specific points or sections in the audio, with very few lines of code. Importantly it also protected us from any cross-browser issues and allowed our designers to effortlessly create a custom skin for the player.
So this was the control, but what about the media? Well this part was a massive team effort. Henrik managed to provide a very accurately timed transcript. We had hoped to use the subtitles in SRT format but for convenience we parsed them or rather Scott Downe parsed them into JSON format.
One of the bigger issues we encountered was that we only had the transcript in English and the timings for the Danish transcript were naturally different. Luckily we had accurately timed Danish subtitles and legendary Bobby Richter on hand to convert the subtitles to individual words complete with their timings, which he did by cunningly interpolating the timing of words (based on word length) and based on their in-subtitle position. All knocked out in about 10 minutes and in 20 lines of code. It worked surprisingly well, of course you need to be able to understand Danish to truly tell. We could have probably parsed the subtitles into the transcript on the fly but due to time limitations we made them static.
Perhaps an aside not directly related to audio, I managed to hack together some code that allowed highlighted transcript text to be placed in the ‘share’ box, and grab the timings of the first and last words, from there it was pretty much straightforward to make this excerpt tweetable.
This whole endeavor was very much a group effort, a huge thanks to the popcorn.js team, who made joining their IRC feel like walking into a pub full of friends.
Special credit and thanks then should go to Scott Downe, Bobby Richter, Barry Threw, David Humphrey, Brett Gaylor, Ben Moskowitz, Christian Valentiner, Silvia Benvenuti and of course Henrik ‘Tank’ Moltke whose baby all this was. It was great being part of such a talented team. Awesomesauce indeed.
- The Hyperaudio Pad – Next Steps and Media Literacy
- Breaking Out – The Making Of
- Breaking Out – Web Audio and Perceptive Media
- Altrepreneurial vs Entrepreneurial and Why I am going to Work with Al Jazeera
- HTML5 Audio APIs – How Low can we Go?
- Hyperaudio at the Mozilla Festival
- The Hyperaudio Pad – a Software Product Proposal
- Introducing the Hyperaudio Pad (working title)
- Accessibility, Community and Simplicity
- Build First, Ask Questions Later
- Further Experimentation with Hyper Audio
- Hyper Audio – A New Way to Interact
- P2P Web Apps – Brace yourselves, everything is about to change
- A few HTML5 questions that need answering
- Drumbeat Demo – HTML5 Audio Text Sync
Add new tag