My degree has come to an end and the show has gone down really well. Sol is getting a great amount of online attention and interest from a few individuals.

As it stands, the model does very little so I had to make a demo video that shows how it will eventually work. The video is a complex series of button presses done in time with a pre-prepared track and LED animations. I insisted on doing this in one take which was a tedious but rewarding pursuit!

The next phase of this project will be to develop the software side of things in PureData and have Sol communicating with it over USB. This will allow me to refine the functions and workflow and get a few units out for testing. If all goes well I hope to eventually have the device working independently, without any external hardware. By incorporating serial communication over USB Sol could also be used as a controller for production software or digital DJing.


I wanted a rubberised finish on my whole model, similar to Aiaiai’s TMAs. This proved quite expensive and difficult to achieve without using industrial techniques so I set about mimicking the finish using matt paint and different sanding methods. I also built rigs that allowed me to move and rotate the model pieces while they were wet.

I decided to build my prototype by hand from a low density aerated plastic called Cibatool. It has similar properties to wood but has a uniform density and machines really nicely. During the build I became well acquainted with a lovely hand operated mill and rotating chuck. If I was to have one machine in my future workshop, this would be the one.

I’ll miss this old thing…

In order to get through the daunting task of soldering 64 LEDs I set up a system, making sure that everything was well thought out to the end. After all, a single mistake could mean 63 subsequent mistakes! Using perf board meant I could have multiple connections on one LED leg. Once prepared, I inserted the LED modules into a rig and wired them up.

I was surprised at how beautiful the result was…

So far I have been doing all my testing with high output red LEDs simply because I had them lying around after my monome build. Today I did some quick tests with other colours and realised the impact that this decision will have on my final product.

Aspects such as wavelength, intensity and viewing angle are all important factors when choosing LEDs. I had a similar issue with my monome build but the forums had plenty of discussion on the topic…now I’m all on my own! At the moment I’m leaning towards amber/orange for its warmth and high output per Watt. Light around the red/orange portion of the visual spectrum also happens to be optimum for visibility in low-light conditions (hence its use in street lamps).

Meet MAX. He’s very small but VERY useful. He will allow me to control 64 LEDs with only 3 pins on my Arduino. However, I thought he was too cute to perform such a big task so I nastied him up a bit using the worst soldering iron I have ever seen.

Now he looks like a mini deforesting machine…



I did some tests for a purely aesthetic feature of my device. The light from the upward facing LEDs will ‘bleed’ into a surrounding perspex ring. This produces an interesting visual feature for any audience that may be watching a performance. The thinner perspex (1mm) produced more seamless transitions between the LEDs with less defined illumination. The frosting also distributed the light a bit better. It is also worth mentioning that the LEDs in the videos are pointing directly at the Perspex strip. In reality the light will come from the side of the bulbs…more testing!