CHI 2013 BBC interview

The interview I did about the Music room at CHI2013 is now online

[here] 

Wow I talk fast... 

lost weekend

I've been writing a student exercise and began thinking about one writing a single user adventure now titled interactive fiction. The idear was to use hashmaps to represent each room, these would be read in from an xml file.

The things I wanted to do was create an audio game - use speech synthesis to get the text to be read out so you could play it on an iphone or something with your vision else ware.

I was chatting with Ruth and we both agreed the narrative and characterisation and mostly dialog in games is in my experience quite pitiful. So the notion would be to bring this to the centre of the game so you would focus your efforts on it.

So one thing irritates me about interactive fiction it all appears to be written in the first person "You see this, you see that". I thought it would be good do something in the third person ( Lara picks this up, Lara looked round the room to do this).

I think the second thing I liked the look of was introducing a companion who would travel with the main character to get more description out of her/him. In the case of doing a tombraider like interactive fiction I like the notion that someone goes with Lara and they don't get on.

"You want me to jump of this ? are you crazy, do I look insane to you or something".

Looking at the sites on how to write interactive fiction this appears to be a no no - it is thought to interfere with the game begin run by the player.

So looking around and after an excellent wikipedia article on audio games, strong disabled access notion. I'm not against this but I think with mobile audio systems a strongly audio game would be good. Most of the audio games require good use of live stereo sound to get positions.

I checked out the iTunes appstore and after a copy of advent (ure) there is nothing strongly audio game is ( lots of audio in games but nothing audio centric).

http://www.audiogames.net/page.php?pagefile=articles

Ruth and I thought it would be fun to have an augmented reality game where you operated like an adventure but you do this in the real world. So you create an alternative London but set in the spaces of the real one. Naturally you can still walk into any public part but inorder to go from one space to another in the adventure game and get an audio description you have to solve a puzzle(unlock a door, carry an item or something). I quite like this. I guess to be sure of it being interesting I would have to look at a lot of similar augmented reality games. The HCI lit is full of them for some reason.

I should really get on with some proper research but I might come back to this at a later date. No in the end i realised this was too complex for the M255 students to handle.

Big Du

I've written before about audio interfaces now. Apple have produced the new shuffle with an audio interface ( primitive but its prime time). At first I wonderd that they had done speach syntetis on the pod its self ( nice ). Looking at the advert and I realise that the magic is all done on ITunes ( smart) but this does mean that Apple have released the first really big upgrade to the voices on the mac for a long time.

I'll try running Alex on some text to see what its like.

Lets see what this will lead to. I like the idea that we will see more audio interfaces on the other Ipods.

Podcasting experiments

I was reading a student essay on podcasting and while I was had a number of thoughts for experiments. 

1. Use of enhanced sound in podcasting. 

How much do students remember in a podcast? does the kind of information they retain change in a podcast vs written word. It is asserted (Shlosser & Burmeiser/Warbutern) that audio leads to greader understanding and clarification of concepts but with a reduction in 'facts' retention compared written text.  

Pod casts it is said can carry over some of the 'feel' and atmosphere that pure text cannot. 

Experiment: Create two identical podcasts (content wise) first one is a straight audio content ( speeach) the second one includes more 'atmosphere' (sound effects, use of imagery?) . Compare retention.

2. Blended education (Porter 2006) this asserts students need to pull information together and that modern learning methods imply the student is doing more than passive listening. 

Experiment: To identical podcasts but one keeps asking the student to do something ( look up in wiki or follow the podcast and type in what is said). She which has best retention.

3. Isolation ( Lee/Chan 07). The use of voice is suggestive that students form a stronger link to teacher and so less isolated and less likly to drop out. 

Experiment: Two pod casts again Use both enthusiastic and dull voice to see which students are more likely to stop listening too and when.  

University of Sheffield - Steve Whittaker

University of Sheffield - Steve Whittaker

Intreasting lectore by Steve Whittaker on memory. He says that most of the work done shows that while people might spend a lot of time saving memories they don't use them that oftern. For exampe me metioned research the 43% of book marks are not revisited. or that most corporatations show little document reuse and that having recordings of lectures don't help students scores.

Intreasting. Did show off cool device that records sound and lets you take notes. The idea was that the notes index the sounds and so clicking on an old note jumps the audio stream to that point ( when you made the note). I always wanted to do something similar with a button on a recording ipod. That is it always records but marking makes the last 5 mins stored.

Also showed cool zooming text tech. Basically the way to get significant words is to look for words occuring more oftern than it does in general lanuage. Does produce some quite cool results.

Mentioned website where people have labled significant locations ( something to do with BBC and first memories... I can't remember).

Fourth experiment just completed makes us 30% of the way to statisitcal salvation.

Still the memory stuff makes you wonder why you bother to blog at all ?

Use for spatial audio idear

I was talking to Paul about the ipod shuffle and he mentioned that he thought that it was hard to have a screen less interface. I quite agree with him and said I wondered why. I have always been intreasted in try to produce an audio interface but would like to get beyond the phone interface of ‘press one for action a press 2 for action b. wait for a long slow list to get to the option you want’. With icons, location and text you can be reminded and skip to the option you want pretty quickly. The subject moved on to Simon’s excellent audio interfance and we came up with the idea of multiple speakers all talking at once.

Imagine you have a pair of head phones, you have a key pad with say five buttons and the machine wants you make a choice from. The idear is that all five voices sit speaking out aloud. Each voice gets its own speech position in space and you use your listening in party auditory system to identify the one you want. So imagine one voice repeating ‘option one to revise your document. option one to revise your document.’
Another voice repeating
‘Option two to create a new document. Option two to create a new document….’
And so on for all five voices.

The spatial audio (position in stereo field) would help which of the appropriate buttons to click/press. The distance to speaker plus variation in speakers (male, female, old, young, accents) would act both as further ways to separate out the voices (perhaps).
Clearly there must be some natural limit to the number of voices you can differentiate but perhaps 4 five would be useful. Worth testing out.