Hi I am artist working with sound, sculpture and electronics.  I am currently doing a PhD at Oxford Brookes and writing up some of the experiments and ideas I am working on here.

My research lies between, sound, computing and music.  I am exploring ways of communicating to machines through simple sound and rhythm.

As part of this I am developing software that slows down computer communication systems so that people can interact with computers in the same way machines communicate.

There is quite a mix of stuff here from workshop documentation, to software and hardware development, my experiments with sound and rhythm and theoretical stuff.

You can see more of my work at…



Probing the pHAT DAC

I am making a new piece in which a sine wave will be clapped as a binary rhythm and converted in real time into an audio wave form.

We had a few of the pHAT DACs for Raspberry Pi in the studio so I thought I would use one to try and get an idea of how they deal digital information.

I used a Raspberry Pi 3, put on the DAC ran the config file from the link below and it worked pretty much straight away.

Then I installed Audacity on the Pi, made a 16 bit 44.1 khz Sine Wave and played it out through the DAC.


I then soldered a few connectors onto the DAC so that I could run it into my logic tester and see what data is being outputted by the Pi before it is converted by the DAC.


Then I spent ages trying to work out what was going on with the digital output.


I still need to do a bit of research but it looks like for each sample it sent is encoded as 2 bytes of information (16bits) in a frame that is 32 bits long. I don’t really know why that would be, but from nosing around on some forums I saw this quote.

The frame is always 32 bits per sample, and there are two samples per frame, so you need 64 clocks to transfer a frame. So BCLK is always 64 times LRCLK. With 48 kHz sampling, BCLK is 3.072 MHz.

With 256x oversampling (modulator clock MCLK at Fs * 256), BCLK is MCLK / 4.

I found that here… and will hopefully get to grips with how DACS work soon.

Learning Steve Reich’s Clapping Music

I teach on a Fine Art course at Liverpool Hope University and one of my colleagues is Tony Smith is a great drummer as well as a painter. We are always chatting about drumming and have decided to have a stab at Steve Reich’s Clapping Music.

Its a beautiful piece that is simple in someways 12 bars with one clave that shifts in each bar. But obviously it becomes rhythmically very complex and is a bit tricky to learn. So we are doing one bar a week and will perform it next term.

Will add some video soon




Arpeggio Syncopations

I confess to having an addiction to studying flamenco guitar. One of the recurring features that fascinates me is the way phrases are syncopated, for example you may get an arpeggio in cycles of 3 over a count of 4 so the accented notes get displaced.

Its quite hard to explain and I will do a better job another time but here is an example of such a thing taken from a Paco de Lucia solea but played as slow as my metronome goes.

Here is a quick tab of what is going on. The syncopation I mentioned starts in the 2nd bar.



Στοιχεῖα is a dual Euclidean Sequencer. It algorithmically generates rhythmic sequences from an input trigger signal.

The name (Στοιχεῖα, Stoicheia, or Elements) comes from the title of Euclid’s book Elements in which the algorithm was first described.

Txalaparta workshop

Wikipedias article on the Txalaparta starts with a description of it as an instrument for communication. So when the opportunity for a workshop with Enrike Hurtado at Live Interfaces came up I had to do it. My mind immediately wondered what would people communicate about the cider, how were the messages encoded.

Rather than describe this instrument and Enriques work, here is the text that advertised the workshop.

“The Txalaparta is a Basque musical tradition where two people make music by hitting planks of wood with large sticks, communicating with each other according to a tacit algorithm. Enrike Hurtado has developed software which can play the Txalaparta with a human.”

From Enrikes presentation it was clear that as with many folk other traditions, the history of this percussive art form is oblique. The notion that it was an instrument for communication is not well documented and there is even debate as to whether the Txalaparta refers to the instrument or to the way the two performers exchange rhythms.

It was fascinating to hear from Enrike how many aspects of the art form are open to change. For example the planks can vary any size and are not specifically tuned. One of the most notable features of the Txalaparta is that it is never played solo, phrases are played by one player and then repeated or modified by the other player. It functions a little bit like a game.

Its connection to cider seems to be as much due to with the process of making cider as to do with sending messages. From images of pre mechanical cider production its easy to see how the the artform could have evolved.


It was notable how the rhythms had generally become more quantised over the  during

Euba has called xylophonisation

“Considering the historical evolution of the txalaparta – in particular the current ‘xylophonisation’ process where pitch has been added and the rhythm becomes quantised – it is interesting that the digital txalaparta, where the practice is translated into the digital domain, is closer to the origins of the txalaparta in operating with fluid rhythms and non-metric bars, both in its internal algorithms and graphical notation”

Although we did not learn about the structure of the language, as it is doubtful we have

The scores

The key figures who made it more popular

The study of the scores

The gradual quantisation of the form

And then we move to Enriques Software

Rupak Tala

At Flamenco Tours Tarifa this Summer the percussionist Ruven Ruppik gave a talk about studying Tabla. Apparently the traditional way is to study bols for one year. The one he showed was called Rupak Tala and was interesting as it is in a group of 7 beats which are subdivided into 3 groups as follows.

Tin Tin Na | Dhin Na | Dhin Na

This vocal phrasing is then sung with the hands doing a particular counting. Once the tempo is established with the hands and voice. The bol can be doubled, halved, done as triplets and then off the beat creating a dizzying amount of permatations.

Here is a video of me practicing.



Charles Hayward’s, The Bell Agency

This workshop comprises a simple game structure for 3 or more players, The Bell Agency opens out into a mesmerising slow motion sound world. Somewhere between performance and unfolding, it depends on cluster maths, hive mind and the unique details to be witnessed. Open to all those aged from 5 to 6 onwards and willing to participate in this zen, engaging and startlingly educational audial activity.


Sending data through a paper cup

Whilst developing ideas for SPACE’s Schools out for Summer program I was thinking of a way of rhythmically sending information without using clapping or the voice but something else that was physical, lo-fi and cheap to do. Tony reminded me of an old cup and string toy that makes clicks when you pull your fingers along a knotted string.

So we sat around in the Owl Project studio wondering if it could be used to transmit information. I went away and figured out the code for it to work. I ended up writing some software that looks for a start and stop bit and then checks between for 1s or 0s. Have a look at the video below and it should make sense.

So for each letter you needed to tie or not tie ten knots like below. The only way I could make it work was to include a start and stop bit as 1. So an ASCII character is made up of 8 bits plus a start and stop bit. A total of 10 bits.


I found it pretty exciting to print a letter on a screen with such a system. In the workshop situation it kind of worked but it was pretty fiddly for people to tie knots and get it right and then it is actually pretty hard to pull your fingers along the string at the right speed to make it work. So overall I would say it was a little too problematic for a workshop setting.

I was chatting to Alex McLean about the experiment and he told me about the Pre-Columbian information storage system of Quipi.


Quipu with string cords wrapped in different colored threads. Cotton and camelid fibers. Wari Empire (AD 500 – 1000), probably, pre-Inka quipu. American Museum of Natural History, N° 41.2/7679. Photo, courtesy of the AMNH

Alex told me about a show he had worked on at the Open Data Institute

In this show Dave Griffiths and Julian Rohrhuber worked on a piece called ‘Inca Telefax. Listening to Precolumbian Administration without understanding a word’, which is a computer-generated sound installation creating an aural insight into quipu, identifying rhythmic structures.

David Griffiths has written a long blog post about the project here.