The idea behind Buncefield Records and Tapes is to be able to ‘cut’ photographic images to vinyl discs, by translating them to audio tracks. However, rather than many other projects which experiment with converting images to sound to be heard as ‘music’, the ideas behind Buncefield are to do with both ‘musicality’ and data storage – it is important to be able to play the tracks back and recompose the images. In the 1980s I used to load computer programs for the ZX81 and BBC Micro from flexidiscs and cassette tapes – as well as ‘download’ such data from the radio (the BBC transmitted a chess program, I think) using the Kansas City Standard. Those electronic sounds are stored deep in my subconscious, sounds from my childhood. Obviously, we are talking here about two different types of memory.

Audio encoding of images has been deployed by Aphex Twin using spectrum analysis, (and others) and in the world of amateur radio, SSTV represents a well known technique for encoding still images as audio for user transmission. (Check out Daniela de Paulis’ wonderful repurposing of SSTV to bounce images off the surface of the moon!) For the Buncefield project, the most significant precedent is the encoding of images to disc by NASA as part of the Voyager  expeditions beginning in 1977, which left the solar system in Autumn 2013, carrying gold discs which contained colour and monochrome images of earth encoded into their grooves. The discs contain visual instructions about how to decode them, for anyone who finds them. Extraterrestrial DJs, maybe.

First experiments for the Buncefield label involved trying to make something ‘musical’ by using sine-wave bursts. Each pixel of a JPEG was broken down into R,G and B values and encoded sequentially as 10 cycle bursts. This was the technique used in the exhibition Electrofringe – and while the most ‘musical’ method so far, it was rather slow: a 200 x 200 pixel image took 10 minutes plus to play out. (I don’t mind slow –  the relationships between the speed of digital and analogue carrier media is also part of the project – see Radio Buncefield / Songs for Ethylene Glycol)

The procedures for encoding and decoding were built using Processing and a library called ESS.

Around the same time, I experimented with using blob detection and shape recombination in Processing, encoding the x and y axes of shape vertices as rising and falling tones, effectively ‘drawing’ the image. High tones represented the X axes, and low tones the Ys. This made for the prettiest sound so far – some great harmonics and drones. Reviving this method is in the ‘to do’ box.

The method which worked best turned out eventually to be simply breaking the images into individual horizontal RGB lines (first red, then green, then blue), converting the lines directly to audio signals, then inserting a series of ‘flyback’ signals to enable the decoder to identify each one. The images below show screenshots of the analyser sketches. The red bars are the flyback signals, the black traces are the data. The funny looking thing on the right is a rotating oscilloscope I built in Processing for calibrating the signals.

signal anayser shot

Screen shot 2013-02-13 at 8.17.03 PM

And here is a more ‘old school’ image of an older version of the code.

Screen shot 2012-10-29 at 11.17.03 AM

Dub Studio: cutting the discs


(click on the link)

The discs are cut at Dub Studio in Bristol. Lathe master Henry uses a technique he developed himself which enables the more accurate cutting of square waves to vinyl.  Not to give away any secrets here, but it produces a cleaner signal and better data transfer. The discs are cut as one-offs to everlasting vinyl.

Musical disclaimer:

Various listeners have commented on the ugly, grating unmusicality of the work. Though it would be the holy grail to find a medium that is both easy on the ear and holds vast quantities of data, the concept of data storage has to come first, as the key to the project’s success is what you ‘do’ with the product, and sitting in an armchair consuming this stuff as if it were ‘music’ was never the intention.  It is meant to be performative.

Part of this is an attempt to rediscover a rapidly receding sonic landscape – in the future, musicians may well try to sample fax signals*, radio interference, as they have done in the past – but if no-one remembers what they actually sound like, would the point be missed? The sounds Buncefield produces are intended to be source, not interpretation – they are hopefully the kinds of sound that end up getting sampled, it is up to the listener to find them musical.


*play this to the end.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

the sound of data, the noise of photography

%d bloggers like this: