DOI: https://doi.org/10.33008/IJCMR.2019.09 | Issue 1 | March 2019
Author: Stephen Cornford, Oxford Brookes University
Saturation Trails is the collective title of a body of work conducted with the Archaeologies of Media and Technology research group at Winchester School of Art. The work investigates the materiality and operation of the digital image sensor, the ubiquitous photosensitive chip behind the lens of all phones, tablets, laptops, and cameras. To do this, I designed three experiments which would reveal the technics of the photosensitive semiconductor through which all our images are transduced. Image sensors were exposed using methodologies appropriated from the optoelectronics industry in which this component was developed, tested, and manufactured. The three techniques used were an infra-red laser, hydrofluoric acid and X-ray radiation.
Saturation Trails is the collective title of a body of work which investigates the materiality and operation of the digital image sensor, the ubiquitous photosensitive chip behind the lens of all phones, tablets, laptops, and cameras. Historically speaking, artists have not only been early adopters of new image technologies but also first to push these technologies beyond their intended limits. As with Man Ray and Moholy Nagy in the early celluloid era, so too with Nam June Paik and Steina and Woody Vasulka with the arrival of video: ‘investigations into the material support are an integral part of the artistic process’ (Knowles, 2019). However, since the arrival of the digital camera, the technics of image capture have largely remained as opaque to the visual arts as they are to consumers. The work aims, therefore, to redress this absence of practices by exploring digital camera hardware, in turn providing a precedent for an experimental vocabulary of video practices based on the materiality of the image sensor.
To do this, I designed three experiments which would reveal the technics of the photosensitive semiconductor through which all our images are transduced. Image sensors were exposed using methodologies appropriated from the optoelectronics industry in which this component was developed, tested, and manufactured. The three techniques used were an infra-red laser, hydrofluoric acid and X-ray radiation. As these processes, and the facilities used to conduct them, were located in research science facilities, I also adopted the use of the scientific term ‘assay’ to frame them. Here I will discuss the hydrofluoric acid assay, which was conducted at the Clean Room facility of the Optoelectronic Research Centre at the University of Southampton on three days spread over an eighteen-month period.
A ‘saturation trail’ is a technical term used in digital imaging to refer to one of the possible effects of overexposure in a digital image. If a pixel is over-exposed, excess photons will flood neighbouring pixels, usually up or down that column of the image sensor, creating a bright trail. A saturation trail reveals the architecture of the image, or as Susan Schuppli describes of similar video artefacts, is a moment when ‘sensation emerges out of the technical reorganisation of the image-event … out of its material depths rather than out of its mimetic regime’ (2014: 280). Although the artefacts in this assay are not caused by excessive light, but by physical damage to the gridded lattice separating the pixels, the acid etches similar trails, revealing the microelectronic channels in the sensor’s substrate through which the image is readout. The value of Saturation Trails as a title, however, is also metaphorical. How else might we understand this phenomenon of a saturation trail in the context of digital electronics? What are the trails caused by the saturation of our social, economic and media cultures with digital technologies in general and cameras and images in particular? In this statement I propose an ecological reinterpretation of the concept of the saturation trail tied to the materiality of the sensor itself and the acid used to etch it.
The decision to use hydrofluoric acid followed research into the manufacture of image sensors and the recycling of electronics. Acids are crucial to both. This approach eschewed the representative function of the camera and computation, focusing instead on what Jussi Parikka describes as ‘techniques that sustain even the existence of IT – in factories, as well as when discarded electronics are dismantled’ (2015: 89). All image sensors operate by splitting incident photons across a gridded array of pixels. During manufacture this grid is laid down by etching conductive pathways into the chip. Silicon is relatively inert and as hydrofluoric is the only acid with which it reacts it is exclusively used for this purpose. Acids also play a central role in the recycling of consumer electronics, used to leach precious metals from circuits and components by e-waste salvage workers worldwide, a process explored by artists Jonathan Kemp in his Crystal World project (2013). Nitric Acid is used to dissolve base metals such as copper, lead, tin and nickel, which otherwise inhibit the precipitation of gold, the latter can then be recovered electrolytically in a bath of sulphuric acid and subsequently made soluble in Aqua Regia, a combination of hydrochloric and nitric acids (Kemp, 2013). The use of acid in this assay was therefore conceived of as both reflecting the conditions of the sensor’s manufacture and recording in advance their inevitable future dissolution. The acid assay can then be understood as folding the extended temporality of the sensor onto its operative surface: enabling it to record in the present the conditions of its invention, fabrication and degradation (Latour, 2002: 249).
This temporal fold emphasises the mineralogical and metallurgical continuity between rock and conductive circuitry which stretches far beyond the useful lifetime of the silicon sensor as a transducer of light. The temporalities folded together are so conflicting as to be incompatible. The processing speed of an image sensor must operate faster than human perception in order to maintain the illusion of a smooth video feed, reading out several million individual pixel-values in the interval of a single frame. Meanwhile, the design, manufacture, and sale of sensors with ever higher resolutions and faster readout rates is required to maintain the illusion of perpetual growth in both image quality and share-value. The limit speeds of pixels, data and financial markets are known, regularly tested and frequently accelerated, but what is the limit speed of the processes required to dispose of our devices, dissolve their metals and decontaminate the many sites of their production? The microtemporal now of computation and the near-instant representation enabled by the digital camera both materialise in and through metals whose geological present exceeds human history.
The abundance of silicon in the geological strata (where it is second only to oxygen) has now become echoed in the abundance of silicon semiconductors in the technological strata assembled on and above the earth’s crust. As Jennifer Gabrys (2011) has pointed out, the overproduction now common to contemporary consumer electronics was pioneered by Fairchild Semiconductor in the 1960s as a deliberate strategy to make electronic products reliant on these chips. Increasing the volume of chips they fabricated and cutting the price so that they were often sold at a loss enabled Fairchild to ‘saturate markets with microchips in order to allow an emerging technology to take hold’ (Gabrys, 2011: 31). The current prevalence of semiconductors in electronic engineering can therefore be understood as a direct result of a commercial model predicated on market saturation. Following the invention of the charge-coupled device (CCD) at Bell Laboratories in 1969, Fairchild were also the first company to manufacture a commercial CCD image sensor, and were immediately contracted by the US military. Nowadays markets are routinely so swamped with new camera models that spare parts can be purchased inexpensively in wholesale quantities, often years after most users have upgraded. The cameras used in this assay, for example, were selected primarily for the availability of spare sensors on Aliexpress rather than according to their technical specifications.
When the market for spares and refurbishment dries up, the silicon laced products which swamped our distribution centres moves from the field of commerce back to the fields, rivers, oceans and outskirts, becoming laced with traces of the acids used to scratch out a livelihood from the cast-offs of those more prosperous, or as the Basel Action Network put it: ‘the effluent of the affluent’ (2018: 2). In Guiyu, China, for example, the acid stripping of semiconductors has led to:
the routine dumping of aqua regia process sludges that blackened the river banks with the resinous material … A quick test using pH paper on the saturated ground surrounding the tubs measured at a pH level of 0 – the strongest level of acidity (Basel Action Network, 2018: 23).
The isolation and reconstitution of base metals from the geological strata into a functional camera component is achieved by an enormous exertion of energy, labour and raw materials only for its resolution to be surpassed within a year. At both the sites of manufacture and disassembly, the long-term ecological impacts of microchip production and dissolution will outlast the workers whose lives are endangered by their daily exposure to the toxic materialities underlying the production of the digital image. This routine yet rampant ecological devastation is the saturation trail of the electronics industry, the overexposure of the geological surface and low-waged global workforce to logics of extraction, exploitation and contamination that seem rarely to be questioned, what Lisa Parks describes as ‘a dark side of the global digital economy so often eclipsed by bright antiseptic visions of the clean room’ (2007: 39).
In this assay the toxic materialities found inside such Clean Room facilities are turned back against the camera apparatus that relies on them during manufacture. A process appropriated from optoelectronics is repurposed for semiconductor sabotage, or as Gustav Metzger wrote: ‘we shall use science to destroy “science”’ (Metzger, 1965). The acids seen here spreading across the sensor’s surface visualise the damage wrought on a geological scale by the endless production of technologies embedded with semiconductors. The dissolution of the machine-eye visible in each of these videos is enabled by an industrial scale dissolution of landscapes that remains largely hidden from lenses and sensors.
Basel Action Network, Exporting Harm: The High-Tech Trashing of Asia. 2018. Available at: http://archive.ban.org/E-waste/technotrashfinalcomp.pdf (accessed December 6 2018).
Gabrys, J (2011) Digital Rubbish: A Natural History of Electronics. Ann Arbor: University of Michigan Press.
Kemp, J (2013) The Crystal World: Executing a New Media Materialism. PhD Dissertation: University of Westminster.
Knowles, K (2019) Film in Transition: Obsolescence, Materiality and Experimental Cinema. London: Palgrave Macmillan.
Latour, B (2002) Morality and Technology: The End of the Means. Theory, Culture & Society 19 (5/6): 240-262.
Metzger, G (1965) The Chemical Revolution in Art. In: Granta (Cambridge, November 6).
Parikka, J (2015) A Geology of Media. Minneapolis: University of Minnesota Press.
Parks, L (2007) Falling Apart: Electronics Salvaging and the Global Media Economy. In: Residual Media, edited by Charles R Acland, 37-51. Minneapolis: University of Minnesota Press.
Schuppli, S (2014) Entering Evidence. In: Forensis: The Architecture of Public Truth, edited by Forensic Architecture, 279-316. Berlin: Sternberg Press.