I spent a lot of time over the last couple of weeks trying to refine some implementations of my ideas for this project.

I started to modify an old fashioned telephone, that will operate as the main source of interaction for the participant. The fabrication component of this is documented here.

After some trial and error, I managed to get the telephone speaking with my audio interface by cracking open a microphone and appropriating the dynamic microphone inside.

Instead of spending a (vague) estimate of 1/2 to a day wiring the telephone ‘bells’ to the Arduino, I have stashed a couple of speakers inside the Telephone.

I am in the process of wiring up the Telephone receiver switch to the Arduino’s digital input. This switch will keep track of where the phone receiver is. I will need to figure out a way to wire up the pull-down resistor without the space for a full breadboard.

I am also working on setting up a force sensor to communicate via the XBee microcontroller.

This was assisted with the help of Tom Igoes’ ‘Making Things Talk’ book (both PHP and Node.js versions).

However, as I am using Touch Designer as the brains of my operation, I have needed to re-write the parsing code in python.

I have posted the code here.

It seems to be only partially-working at this stage - every 2nd or 3rd value responds to force. The values are not expressing ‘range’ properly either, it seems they are almost ‘on and off’ at this stage.

I also spent time developing the audio-reactive visuals for the installation. When the participant talks into the headset, they will hear a looping soundscape of their own voice in the receiver. A combination of the raw voice sound, and the soundscape are analysed to influence the visuals.

This is a preview of where the visuals are at. I’m not quite happy with them yet. The first part, before it gets too trippy is a representation of the user approaching the installation. The second part is demonstrating what happens visually when they sit down, and answer the telephone.

Hard to see from the video (Couldn’t figure out how to screencap with audio) but it has been really challenging to get some ‘satisfying’ audio-visual effects. I’m still feeling like my method of audio analysis is a little clumsy and could use refinement.

Here is an earlier video of some testing done projecting onto the screens. In this video there are no custom projection screens being fitted.

CONCLUSION

There is more to do. I feel confident I can finish the basic physical components before user testing. However, I’d love to introduce some ‘kinetic sculpture’ to the ‘pile of junk’ that would respond to the voice. I feel this might bring it to life.

I also need to add some sensor interaction that ‘lures’ the user in, I am thinking of using some depth sensors with some additional lights placed on the ‘junk pile’ and perhaps on the chair.

I still need to keep pushing the interaction and the aesthetic to create something more entertaining and interesting.