The #glitch machine B46

The #glitch machine B46

“The #glitch machine is an interactive new media installation that offers participants/viewers to engage with the history of the moving image.

The aim of the installation is to offer new perspectives on the impact, the nature and the influence of the moving image over the last 110 years via a derive through the iconic and influential.

From the Lumiere Brothers through mass media consumption and new media art, this is an opportunity to engage with and re-examine the influence of moving image practice and theory.”



In my very first imaginings of how I might want to develop my practice through this mode of study I envisaged a crank that when turned would activate an algorithm that simulated the effects of the Mandlebrot set on a users personal narrative text.

Sketch of the “Mandlebrot Crank”

In this installation the user would use the crank to manufacture a pictorial representation of the content of the text using the text characters.

The key to this installation, based on my own reflection is centered around the nature of feedback. That is how the user controls the crank to produce a representation of the text that he thinks appropriate to the meaning of the text itself and subsequently how that meaning translates and feeds back into the original text and becomes organic and self-perpetuating.

The specific use of the Mandlebrot equations was linked to chaos theory and how, with enough iterations the seemingly chaotic patterns output from the machine would begin to order themselves in line with natural forms.

I was confident in my ability to build the software needed for this machine using arduino and processing being familiar with the applications of using actuators to drive software systems however, at this point I was unfamiliar with the nature and provenance of this idea in a new media context.

It was more than a year later when considering the content of my production project that the concept of the crank reappeared in my thinking after considering and investigating projects based on projection mapping, glitch and liminality.

Conceptual Layers

Glitch and ideas of liminality, the unfinished, the becoming and the ever developing had began to influence my thinking around the nature of interactive new media art and the cyclical motion of the crank seemed to also embody references to the provenance of my work.

Introducing an element of glitch art into the work also seemed to have echoes of the Situationist International in that the glitch in the context of my thinking was a forced error in a digital system that could produce a parallel, disruptive and subversive language that questioned our understanding of and our relationship with the moving image.

Just as Debord sought to upload error and juxtaposition to “the spectacle” I felt that the crank driven interface might offer a similar means of offering opportunities for recontextualisation and reconsideration of new media art.

Thus the machine as a concept began to develop with multiple layers of historical references encapsulated in the nature of the window/mirror interface that aimed to disrupt the traditional concepts of the moving image, the digital and our relationships and understanding of these concepts that was liminal in its always moving design.

These conceptual layers became as manifold as complex in terms of their inter-relations as my own conceptual dérive through 20th century thinking and as well as representing the application of some of the critical theory I had been researching, there were eventually other layers that became more personal as I sought to investigate the operational mechanics of the machine.

Hard Hands/Soft Hands

Having settled on the crank as the means of interaction with the machine, one of the first problems I faced was the manufacture of such a piece of hardware and it was at this point that the project became collaborative as I sought advice from someone I knew would be able to advise me on such issues; my father.

He suggested that a communtator, a rotary electronic switch would offer the means to produce an alternating signal I required for the crank to drive a moving image image installation; this alternating signal could be processed to simulate a mechanical projector by advancing the image forward or backward with each individual signal received.

Thus we decided that the best course forward in terms of experimentation and development would be through hacking old electric drills.

The process of repurposing these drills was to reverse their functionality: instead of using electricity to drive the rotary motor we wanted to turn the motor in order to produce an electrical signal that would be read into the software system via an arduino interface.

The next problem I faced was that of direction. It was important to me that the direction of the turning of the motor would have an effect on the image produced (i.e. effect the direciton of playback in the system).

At this point I enlisted the help and advice of several people from the local pub who posessed a range of skills and backgrounds and were able to offer mechanical solutions to this issue via discussions and diagrammatical flow diagrams:

Working out directional concerns at the pub

While I was overcomplicating issues with thoughts of sensors and clutches a friend came up with the idea of using a simple two way switch that would be struck by the handle of the rotor and forced into a position whereby it would be able to generate a signal that could be translated as “forward” or “backward”.

Once these hardware issues had been solved it only remained to route these signals through the software environment in order to build the functionality of the machine.

On reflection, it soon became apparent to me that there was yet another conceptual layer within the development and realisation of the machine; while I was of a generation of the rise of computers and digital systems, I was working and collaborating with those of a generation more familiar with industrial, analogue technology.

It seemed that the machine was also beginning to be representative of the relationship between the industrial and post-industrial worlds; while my technology was small, intricate and in may ways soft, my collaborators understanding of the manipulation of technology was larger, more robust and somehow harder.

These seemed to people who had worked with technology directly, very “hands on” as it were, while my experiences had been “softer”, mediated by the computer screen.

While the difference between these hard and soft hands was apparent to me and again representational in a cyclical way of my application of critical theory it also occured to me that these hard hands and their experience had also been mediated by tools, albeit the tools of a more analogue industrial generation.

Thus the construction of the hardware and indeed the software meant that the machine was now taking on an element of itself being a tool of remediation (in line with new media theories) whilst simultaneously representing the very concept of remediation through its relationship with the industrial.

Thus the machine became a détourned artifact, a kind of readymade, in its hacked, repurposed state that was a window/mirror on the industrial and post-industrial worlds that was aimed at producing a screen mediated détourned artwork that investigated our relationship with the moving image and the critical theories that surround it.

The #glitch machine prototype at Chapter Arts, Cardiff June 2012

In June 2012, I was invited to display the machine at Chapter Arts Centre in Cardiff.

To date the machine was still very much in development with no audio and limited ability in terms of the glitch module and its ability to mix the captured (via webcam) and stock film clips.

As part of the development of the machine to this point I had again involved friends and colleagues to submit work to the machine to be glitched, mixed and recontextualised.

However it soon became apparent to me, especially with regard to the film clips I was using that I may have been falling into Duchamps “trap of taste” as I began assembling clips that were of significance to me and represented what I felt were concerned with movement in moving image.

I selected clips that highlighted movement and it was at this point, somewhat unconciously, I began to put together a database of clips that were experimental in their original context (e.g. Jean Epsteins 1928 “La Chute la Maison de Usher”) as well as being concerned with experiments (e.g. James Whales 1935 “Bride of Frankenstein).

The #glitch machine at Chapter Arts Cardiff Invitiation

The machine at this point consisted of the drills in their naked, hacked form as I wanted to foreground the concepts of the industrial/post-industrial relationship I felt had been at the heart of development to this point.

However it seemed that the abstract nature of the relationship between the hardware and the screen was rather inaccessible and this coupled with the lack of audio (and the ability to control that audio) made for an experience that was perhaps lacking in its ability to engage and recontextualise in the way I had imagined.

Description of the functionality of the machine June 2012

Functionality and Software Development

My experience at Chapter Arts gave me the opportunity over the following months to reflect on the development of the machine and to plan and instigate adaptations and redevelopments that I hoped would realise its potential both in terms of its theoretical basis and the nature of the interaction it presented to the viewer/user.

Firstly I decided to move away from the “naked” drill setup in favour of presenting a less abstract link between the interface and its functionality for the user.

To this end I set about developing a setup that was based on only two drills which were representative of the “camera” and the “projector”.

The drills were hidden inside their respective housings and I felt my concerns around foregrounding the industrial/post-industrial were replaced and enhanced with the addition of this industrial representation of a traditional means of production of moving image works that would become the interface of the new media installation.

The Camera Unit (Hardware): The camera unit was now housed in a box that was aimed at being representative of a traditional camera set up; mounted on a tripod with a lens. The lens in this case was a webcam mounted inside the housing and the drill was now only visible via its protruding handle which, when turned, would capture a series of still images (based on the communtator signal) to be processed within the software environment.

The Projector Unit (Hardware): As with the camera unit the projector unit was now housed in a box that simulated a traditional crank driven projector. Room was made in the housing to accommodate the digital projector and give the impression of a self-contained unit. The communtator in the hidden drill would drive the playback of the images and the addition of a switch inside the unit would dictate the direction of that playback

Audio: The issue of the missing audio from the Chapter prototype was addressed by using the signal flow from both the software and hardware to control the audio playback via MIDI integration with Native Instruments Traktor Unit.

The audio unit presented a soundscape that was based on the original audio from the film database across two audio channels:

Channel 1 would play an audio clip picked at random from the playlist of the audio files connected with the film clips for the duration of the video playback on the screen.

Channel 2 would play an audio clip from the playlist for an amount of time determined at random by the software.

The two channels are overlapped and can be effected by the use of the potentiometers (knobs) mounted on top of the projector module.

Text: I wanted to add an extra layer of functionality to the machine that represented some of the key thoughts in relation to critical theory that had informed development to date and therefore introduced a software element that would read in lines from a text file to be displayed at random across the screen.

AUTO/MANUAL MODE: I was also conscious that there needed to be an element of automation involved in the machines operation so that if the user ended their interaction the machine still had a means of continuing its dérive. Thus the introduction of autoMode where if no interaction occurred for 2 seconds, the machine would continue random playback whereas the reactivation of any of the interactive elements would re-engage manualMode which passed control back to the user.


1. Manual Mode activated with Camera unit: The webcam input is displayed top right of the screen and the communtator activates capture of that area of the screen. Text generator unit is on constantly and offers the chance of the captured images being overlayed with part of the text appearing on the screen.

2. Manual Mode activated with Projector Unit: communtator controls playback of all aspects of the screen; film database, captured images, glitched images and text generator.

3. Auto Mode activated (2 second delay): all aspects of playback are randomised including screen sizes of areas 1-4 and appearance of text generator.

The diagram below outlines more specifics of the software environment: