the digital image

A common argument in photographic theory is that the triumph of the digital image as the contemporary form of photography forces a reevaluation of the traditional assumption of correspondence between the image and some form of reality of which it is said to be an imprint.   The argument is that digital  images that begin their life as binary data and are  driven  by algorithms  cannot be comprehended through the conventional  trinity of representation, the index and the punctum.  A major shift has taken place with the emergence of the networked image.

As a photographer I understand  the digital image to be an evolution from analogue photography: to all intents and purposes a digital image made with a digital camera  is  little different to the one that is made with an analogue camera.  I situate myself in the world in the act of photographing,  and  then I use these  working tools to construct visual representations. The  Sony a7R111 digital camera is an automated,   computational and pre-programmed tool compared to  the entirely manual Leica M 4-P analogue camera that was made in the 1970s.   The trajectory  in digital photography is towards the expensive professional high end. This  means  increased  automation,   a pre-programmed apparatus,  and more and more AI being built into the post processing software in order to  counter the competition from the increasingly sophisticated cameras in  smart phones.

Here is a digital image made with a digital Sony-a7 R111 camera:

Here is the analogue photograph   made with  the all manual  Leica M 4-P analogue camera.  The negative  has been scanned into a digital file and then processed in Lightroom.

The differences between the two technologies within this  logic of representation are minimal  when they are viewed on a computer screen after being edited with Lightroom software.  The object —ie., the quartz  and creek in the two images –is known to us as a representation of the object.  Photography is a process that mediates the world with the agency of light to produce legible images.  

From my perspective as a working photographer the main difference between the two technologies is evolutionary. The digital technology is more convenient to use  and  it offers greater flexibility  for  hand held photograph in low light situations–eg., at dawn.   As a photographer I continue to work within the trinity of representation, the index and the punctum, with both digital and analogue cameras.   However,   I do  realise that the image on the computer screen  made with a digital camera resembles the look of a traditional photograph  because the computational processes are currently designed by the manufacturers  to make these data packages look familiar to those working within the photograhic tradition.

However this not the full story since  in our contemporary digital visual culture  it is computers and not humans are the ones who process, sort, store, archive and distribute images. For a computer a photograph is calculable information, not different from other bits of calculable information (eg., those we refer to  as songs, films and books); information that   continually combines and recombines figures, texts, glitches and numbers by passing electronic signals between the nodal points of the internet. 

In becoming computational the image  partakes in a different kind of logic to the traditional one that is underpinned by representation and indexicality. Daniel Rubinstein, for instance,  argues that the difference between the two technologies–analogue and digital– is metadata. The digital camera produces  descriptive metadata which is mechanically generated from such details as date, location, camera make, owner, keywords, and is carried within the file. The other type of metadata  comes with publishing the image on the web (eg., a blog, Instagram, Flickr) in the form of tags, comments, ratings, number of viewings etc  and it  is stored independently of the image.

This type of online communication  metadata would also apply to a scanned analogue picture when it is published online.  In today’s visual regime an image can be uploaded to someone’s Facebook stream in the morning, “liked” and tagged at various points of the network and by the evening re-emerge as part of diverse and varied series, search results and image-sets that have no linear connection with the event of the original upload: it is trending on twitter, it is siphoned into image mashups, remixed into palimpsets and aggregated with other bits of information to form new images.  Once uploaded online, an image can appear anywhere there is a networked device and it can do so simultaneously across the entire globe. 

By re-writing the image as machine-readable text, metadata facilitates the identification, discovery, retrieval, and dissemination of images online, thereby problematizing the relationship between authors,   readers of images and subject. The computer sees the networked image on Flickr through the packages of metadata linked through the communication networks that  distribute, mediate, assemble and re-assemble electronic data. This digital image traverses the networks not as bounded snapshots,  but as dynamic packages of data. In informational capitalism there is a  perpetual circulation  of re-packaging and re-processing of data  with the constant bifurcation  of the image as packages of data into divergent but interconnected narratives.   The photograph is valued not as a singular object but as a resource to be deployed in endless and variegated successive contexts.

Consequently, the photographic image online has little to do with the traditional semiotics of representation, economy of signs, signifiers  and indices as it has more to do with  packets of data that are  distributed according to certain rules. The turn is  towards the mathematical and the algorithmic  and away from  the art-historical models of conceiving the image in visual terms as an archival or curatorial object.    This is a turn to  the logic of  self-duplication and mutation of a  computational world,  in which the algorithmically processed digital  networked image  is connected to the principles of computer theory, self-referential replication,  and to Gödel’s undecidability theorems, Heisenberg’s uncertainty principle and the paradoxes of Turing machines.