cp2david

[cp2david]

Goldsmiths BSc Digital Arts Computing.

Creative Projects 2.

Gitlab project repo - https://gitlab.doc.gold.ac.uk/droll002/comp-arts-practice-2

Please note: this blog has 3 pages that can be navigated between at the bottom of the page.

[cp2david]

CONDUIT

CONDUIT is an installation work that utilizes 3 different physical states of matter to engage its audience with oil-based materials. The work hopes to promote reflection upon the material history of such ancient materials that are often presented to us in the form of brand-new products.

Until affected by human interaction, CONDUIT is a moment of pause for a slow material that is now being consumed rapidly by the parasitic forces of capitalism. The work in its exhibited form comprises of a black rectangular plastic cattle trough filled with black viscous liquid imitating oil, with 3 clear plastic tubes running from 3 external vacuum pump DC motors into the bottom of the container. The vacuum pumps are attached to an Arduino control circuit that reads force values from 3 corresponding FSRs and maps the values to the speeds of each motor. Depending upon how much or how little force is applied to the FSRs, the motors speed up or slow down, pumping air bubbles through the liquid that rise to the surface. When force is removed, the pumps lie dormant. The work is arranged with the FSRs spread out on a flat surface, perpendicular to the orientation of the cattle trough, with the audience invited to interact with the force sensors by hand or with a series of 3 glasses, each containing a different quantity of sand or water.

CONDUIT has three dynamic elements (only 2 being shown in the exhibition images below). The first is the liquid’s surface – black but reflective, affected by the second dynamic element – pressure, placed upon the surface of the FSRs. This pressure alludes to the timeline of becoming that plastic has gone through. Plastic is ancient biological matter, presented to us in a state of inertia. Initially transformed from biomass into oil through pressure and force from the earth over thousands of years, oil’s volatility is exploited by capitalist processes of extraction, combustion, polymerisation and polycondensation to be burnt away as fuel or used in the manufacture of plastics. The third dynamic element is sound – when pressure is applied to the FSRs and the surface of the oil begins to shift, synthesised bubbling sounds are triggered that decay and fade over time, releasing the bubbles from their liquid form and transporting their texture through the air audibly. In overtly presenting dynamic interaction with oil-based materials, I hope the audience will contemplate where such materials will go next.

[cp2david]

Background Research.

‘Plastic is our kin, it's our relation. It's from ancestors – organic ancestors from a long time ago. And if you neglect your relations to that, then you're bad kin. Even when plastic is misbehaving, which means it's being bad kin, you can still do good kinship with bad kin.’

The genesis of the project came from the article ‘Anti-Colonial Science & the Ubiquity of Plastic’. The article asserts that the consideration of plastic as a ‘synthetic’ material is a direct product of western capitalism - ‘The opposite of nature isn't plastics. That's a very false dichotomy that comes out of colonial science.’[1] I connected this to Foucault’s ‘The History of Sexuality’, in which he describes power as a relation that reinforces and reproduces its own future to create a power relationship.[2] How are our relationships to materials a conduit for capitalist colonial power? How can we reorient ourselves in relation to materials to promote mutualist symbiosis? How could sentiments toward materials be shifted through artwork and artistic practice?

(above) Hayden Dunham is an artist that explores similar questions in her work, creating dynamic, transformational installation works that interact with the body. ‘Dunham’s works embody ideas of transformation and a process of facilitation, where objects are conditioned and supported through their individual internal transformations.’[3] She is primarily concerned with ‘floor-bound works’[4] to make a space for corporeal awareness and heighten attentiveness to the body’s interaction with the materials present. This affectation was something I attempted to emulate in the fabrication stage of my project.

(above) Ryoji Ikeda’s ‘Test Pattern’ is similarly floor-bound, with rapid-moving binary patterns cast downward from projectors above. However, it explores a different intensity to Dunham’s work, creating a deluge of audio and light upon the viewer – condensing, compacting and saturating sensory experience as a reflection of the global present rather than creating a vessel for the audience. The intention of CONDUIT is to operate between these two extremes – referencing the violence of exploitative systems whilst still holding space for the viewer to reflect upon their own agency in relation to the subject matter.

(above) Olafur Eliasson’s ‘In Real Life’ exhibition attempts to warp sensorial perception in order to shift the ways in which the audience engage with their surroundings. ‘Many of his installations play with reflections, inversions, after-images and shifting colours, to challenge the way we navigate and perceive our environments.’[5] In the context of my own work, I am particularly interested in temporal shifts using the senses – ‘after-images’. The audible aspect of CONDUIT attempts to extend the visual dynamics of the work through the air, continuing to interact with the audience’s body after tactile connection between them and the work has ceased.

[1] Liboiron, Max. ‘Anti-Colonial Science & the Ubiquity of Plastic’. Frank. http://www.franknews.us/interviews/206/anti-colonial-science-the-ubiquity-of-plastic.

[2] Foucault, Michel. The History of Sexuality. 1st American ed. New York: Pantheon Books, 1978.

[3] ‘Hayden Dunham - Artist - Andrea Rosen Gallery’. http://m.andrearosengallery.com/artists/hayden-dunham.

[4] I M M A. As Above, So Below: Portals, Visions, Spirits & Mystics, 2017. https://www.youtube.com/watch?v=ZDLAN51fkzQ.

[5] Tate. ‘Olafur Eliasson: In Real Life – Press Release’. Tate. https://www.tate.org.uk/press/press-releases/olafur-eliasson-real-life-0.

[cp2david]

Context, Users, Goals

Due to the interactive and conceptual nature of the work, it’s intent is to be exhibited inside the ‘white cube’ exhibition space. CONDUIT leverages the tradition of the vacuum white-space to invite meditation and criticality upon its subject matter. Therefore, the work is perhaps geared toward an audience interested in critical engagement with art, but offers a more playful, experimental experience for those to which this holds less importance. The work utilizes a dark colour scheme not only to reference the subterrestrial history of the materials used, but also to present itself in contrast to the pure-white, LED-bleached surfaces of the gallery. In an interview, Rick Owens suggests black has ‘nobility’ and ‘reserve’[6] in relation to its inherent modesty recognised by figures such as nuns and preists. The work’s presence should be clinical and modest, but invite intrigue. Due to being unable to exhibit in the gallery space and having little access to fabrication tools, my creative decisions have been impacted. These decisions will be detailed further below.

[6] https://www.youtube.com/watch?v=R0NqydSer5I

[cp2david]

Creative Process.

Physical computing

I began the project with very little knowledge of physical computing, therefore I adopted an exploratory and iterative process in creating the outcome. Upon purchasing an Arduino starter kit, I began to work through basic LED and relay tutorials.

Once getting a relay circuit running with a 5v motor, I purchased a larger 12v DC vacuum pump to prototype with. I chose a 12v motor because of its higher litres per minute air flow rating – I wanted a large and obvious bubbling effect. However, this meant moving onto using an external power supply unit which I found difficult to wire initially. I had to understand the difference between the horizontal power and ground rails, and the vertically oriented circuit area, as well as the schematics for the components in order to wire them so power could flow correctly. After purchasing a 12v battery pack (8 x 1.5v AA batteries) and wiring the 12v motor with the relay, I began to research the ways in which I could map sensor input values to motor speed. I was initially confused, wondering how I could use a 5v PWM signal to power a 12v motor. I came across motor drivers while searching through forums and realised I had a couple included inside the Arduino starter kit. Wiring this was the next challenge – I had to again figure out the functions of each pin on the motor driver chip, and how they corresponded to the digital output pins on the Arduino. I referenced the lastminuteengineers.com guide (https://lastminuteengineers.com/l293d-dc-motor-arduino-tutorial/) for both wiring and programming this. The next step was to wire in an FSR – I used an adafruit guide (https://learn.adafruit.com/force-sensitive-resistor-fsr/using-an-fsr) for this, using a 10Kohm resistor and wiring it into the Arduino’s 5V power rail on the breadboard. I used the same tutorial to figure out how to program the Arduino so it could read and output the FSR input values to the serial monitor. I could then work out how I needed to map each value for motor driver PWM. The FSR input values were between 0 and 1023, so I mapped them to the motor power bounds of 40 to 255 – I used 40 as the minimum map value because I wanted the motors to pump even when little pressure was applied. I then added a threshold of a force of 50 to accommodate for sensor interference – the FSR never read a value of zero and fluctuated around the 20’s when no pressure was applied. I initially ran into problems using pulse width modulation, with the motor behaving strangely. I realised this was because I was using digital output pins that weren’t PWM compatible – I had to switch the analogWrite pins to ones that had a tilda next to them on the board (pins 3, 5, 6, 9, 10, 11). After achieving my desired interface between motor and FSR, the next stage was to scale up by adding two more motors and FSRs. I was initially unsure whether I’d have to use millis() and a control system to ‘multitask’, and referenced these tutorials (https://www.embeddedcomputing.com/application/industrial/motor-control/designing-control-systems-with-multiple-motors and https://learn.adafruit.com/multi-tasking-the-arduino-part-1/ditch-the-delay), however after discovering this (https://youtu.be/dyjo_ggEtVU) youtube tutorial, I realised it could be done without millis or delay. This made the programming much simpler.

[cp2david]

.

[cp2david]

Fabrication

I made several drawings of potential container designs, however due to workshop closures I had to make do with readymade vessels.

Each design I came up with had slightly different thoughts behind it, playing with levels, materials and shapes. For practicality I decided to opt for a plastic cattle trough (final image) – the material, colour and agricultural intention of the object seemed fitting to the concept of the work. I used 6mm external, 4mm internal diameter clear plastic aquarium tubing for the air to travel through, and 4mm diameter air valves to prevent any liquid from being sucked back into the vacuum pump DC motors. I liked the clinical effect of these tubes - their relation to aquatic life also seemed to match my concept.

I also had to decide upon the liquid to fill the container with – it had to be black, viscous and safe to use. I was initially experimenting with cornflower and ink/paint, but during critiques I was given the suggestion of using TV-prop liquids. I came across the compound natrosol hydroxyethylcellulose, which is used within film and TV for gunge. It is water-soluble and non-toxic, so it is safe for an audience and safe for domestic disposal. I bought the compound from eBay pre-coloured in black to mimic crude oil. I had to experiment with the ratio of powder to water and found that 400g of natrosol could be added to 20 litres of water to create my desired consistency. Too thick, and the liquid would not bubble correctly, only making a few larger bubbles – too thin and the liquid wouldn’t appear murky enough.

Beginning assembly -

 Initially I was using the twist and tape method of joining wires, however this proved difficult when many of the wires I was using did not have a stranded core. Many of the connections were intermittent. I decided to solder the FSR to the jumper wires to create a more stable connection.

[cp2david]

.

[cp2david]

Experimenting with natrosol viscosity - blowing bubbles into the bottom of the container each time I changed the ratios of water to natrosol to get closer to the desired consistency.

[cp2david]

.

[cp2david]

Assemblage 1 - floor-bound work.

The first assemblage of the work was on the floor, true to my initial idea – this perhaps would have had a more impactful presence in the white-box gallery space. On my living-room floor, the interactive element and effect of the work was not evident in photos. This is partially down to the vacuum pumps I was using – they did not pump as much air through the liquid as I’d have liked. I wanted a much more vigorous bubbling effect, so perhaps using double the number of motors, or entirely different, higher-powered motors altogether might have rectified this. I didn’t have time to experiment with alternate motors and power levels, so I instead altered the placement of the work.

[cp2david]

Demonstrating the final Arduino circuit and its functionality.

[cp2david]

.

[cp2david]

 Assemblage 2 – Table-top work and addition of weighted glasses.

Bringing the artwork upwards allows the audience to still look down onto it, but allows a greater sense of direct skin-on-artwork tactility. Floor-based pressure sensors meant that there was less of a degree of control over how much the artwork bubbled, since people are more likely to shift their entire weight onto the sensor. Adding the glasses created an additional aesthetic element that I thought the work needed – assemblage 1 was too simple. The glasses, containing sand and water also reference my concept – sand is used in manufacture of both glass and silicon components and so is cohesive with the materials already in the work.

This orientation of the work was displayed at the Third___ exhibition.

[cp2david]

.

[cp2david]

Audio & audio programming

My initial idea was to program a live reverb system with a microphone that hung over the artwork, however this might have proven difficult practically – getting the microphone close enough to the work to generate a sound as well as the reverb system itself. According to this (https://www.kvraudio.com/forum/viewtopic.php?t=551071) article, reverb is difficult to calculate, so I moved on to thinking about applying a convolution reverb to a live signal using Jupiter & Python. However, I could only find examples where convolution was applied to pre-existing mp3 samples. I wanted a time-stretching functionality that could potentially operate in tandem with live input from the Arduino, so I discounted convolution as an option. I moved onto JUCE, an audio programming language that can be used to write plugins such as gain and delay. I realised that synthesising bubble sounds might provide more control over the sounds used, so I researched the techniques of musicians such as SOPHIE that use bubbly textures in music.

Example track:

SOPHIE - LEMONADE - https://soundcloud.com/msmsmsm/sophie-lemonade

I found that utilizing very short delay times[7] as well as chorus and flanger[8] effects could create buoyant textures from raw wave forms in Ableton’s built-in synthesiser, ‘Operator’. The videos below try to display some of my experiments – the bar at the bottom of the screen displays the Operator instrument with the effect chain to the right. Operator creates the raw waveform audio signal and passes it down the chain to be manipulated by each consecutive effect.  

I closely followed tutorials by The Audio Programmer to create the delay plugin – copying between audio buffers & looping the audio signal. I added sliders to manipulate delay gain and delay time so I could calibrate the sound to suit the physical bubbling of the work. In order to interface the plugin with the Arduino, I had to first open the plugin as a VST in Ableton. This was initially obfuscating, as I was using a Windows PC – most JUCE tutorials are on Mac. I had to download the Steinberg VST SDK and alter JUCE’s global paths and plugin formats to save the code as an executable and VST3 file. Once I had got the plugin up and running in the DAW, I had to interface the DAW and the Arduino. Ableton has its own ‘Connection Kit’ plugin that makes use of the standard firmata Arduino code example, however I didn’t know how to merge my FSR and motor sketch with the standard firmata sketch. After finding this (https://youtu.be/BzLKtor-6Vg), I followed the guide, using serial out on the Arduino and loopMIDI and HairlessMIDI on my PC to allow Ableton to respond to input from the FSR. I kept receiving the error ‘Warning: got a status byte when we were expecting 1 more data bytes, sending possibly incomplete MIDI message’ in the Hairless MIDI console – in order to rectify this I had to play with the baud rates on both the Hairless MIDI receiver and on the Arduino. I also had to comment out any serial printing as this seemed to interfere with the MIDI signal.

[7] Elektronauts. ‘SOPHIE’s Sound Design (Digitone or Monomachine?)’, 26 July 2018. https://www.elektronauts.com/t/sophies-sound-design-digitone-or-monomachine/60043.

[8] Resident Advisor. ‘SOPHIE - Bipp / Elle · Single Review ⟋ RA’. https://ra.co/reviews/13209.