CODEC: on Thomas Ruff's JPEGs

This article was downloaded by: [T&F Internal Users], [Carol Alexander]On: 10 March 2014, At: 05:59Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Digital CreativityPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/ndcr20

CODEC: on Thomas Ruff's JPEGsIngrid Hoelzla & Remi Marieb

a School of Creative Media, City University of Hong Kongb Independent writerPublished online: 16 Oct 2013.

To cite this article: Ingrid Hoelzl & Remi Marie (2014) CODEC: on Thomas Ruff's JPEGs, Digital Creativity,25:1, 79-96, DOI: 10.1080/14626268.2013.817434

To link to this article: http://dx.doi.org/10.1080/14626268.2013.817434

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and ourlicensors make no representations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressed in this publicationare the opinions and views of the authors, and are not the views of or endorsed by Taylor &Francis. The accuracy of the Content should not be relied upon and should be independentlyverified with primary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

CODEC: on ThomasRuff’s JPEGsIngrid Hoelzla and Remi Marieb

aSchool of Creative Media, City University of Hong Kong; bIndependentwriter

[email protected]; [email protected]

Abstract

For his JPEGs series started in 2002, German artistThomas Ruff used digital photographs taken byhimself and from the web, compressed using themaximum rate, and then decompressed into large-scaleprints. This method of hypercompression and enlarge-ment exposes the mathematical infrastructure of theJPEG image, i.e. the pixel blocks into which theimage is split during the compression process. In sodoing, Ruff turns a digital artefact (pixilation) into adefault aesthetic, thereby exposing the JPEG astoday’s default mode of viewing images: online andon-screen. Based on a thorough study of JPEG com-pression and its artistic use by Ruff, we show that withthis shift from geometric projection to algorithmic pro-cessing, ‘photographic’ no longer denotes a specificmode of image creation, but rather a specific mode ofimage processing and that the new ‘architecturalorder’ of the image is the mathematical matrix usedduring compression.

Keywords: Thomas Ruff, contemporary photography,JPEG compression, web, internet, archive, archetype,database, algorithm

1 Introduction

This article examines the quantic jump from geo-metric to algorithmic, taking the example ofJPEG compression. We will show that, with theJPEG codec, the projective logic of the image—the projection of a 3-D object on a 2-D planeand the reconstruction of the 3-D image in thehuman brain—is superseded by an algorithmiclogic in which image development and printingare replaced by image processing and display.Here, the photographic is no longer the ‘drawingby light’, but is a generic term for continuoustone images that can be photographs, paintings,drawings, computer graphics or any combinationof these. More precisely, the photographic is ageneric term for a type of aesthetic distributionfor which the compression algorithm works best:in the age of ‘JPEGability’,1 the photographichas become a term for the compressible image.

Thomas Ruff’s JPEGs series (2009) will serveas a starting point for our investigation into theshift from geometric projection to algorithmic pro-cessing for two reasons: (1) its concept is based onan investigation of the aesthetics of JPEG com-pression; and (2) because its material is mostlysourced from the web, the series is also a statementon the new, digital economy of the image charac-terised by an online and on-screen exchange viathe World Wide Web (in addition to email, cellphone and now mobile broadband). As such, theseries is a starting point for our reflection on thenew place of the image, the web and its mode ofaccess, the digital screen. The ‘digital revolution’not only concerns the production, storage and

Digital Creativity, 2014Vol. 25, No. 1, 79–96, http://dx.doi.org/10.1080/14626268.2013.817434

# 2013 Taylor & Francis

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

distribution of images (digital, compressed, viaweb servers), but also their access and mode ofappearance (online, on-screen, in motion). Ruff’sJPEGs series (while confined to still images)will serve to highlight the disruptions in theimage economy fostered by digital technologyand the new relation between web and archive,image and database, archive and archetype, bothof which testify to the new regime of digitalstorage and dissemination: JPEG compression.

2 Thomas Ruff: from the Portraits tothe JPEGs

German photographer and artist Thomas Ruff hasalways been at the forefront of contemporary pho-tography and its reconceptualisation through con-ceptual art practices such as found footage, firstanalogue and later digital. Coming from a docu-mentary background, and training with Berndand Hilla Becher at the Dusseldorf Art Academy,he fought the representationalist notion of

photography, and the idea of photographic trans-parence from early on, propagating the autonomyof the photographic image.

He first became famous with his Portraitsseries, started in 1981 with a series of passport-style colour photographs of friends and col-leagues at the Art Academy. Annoyed with thereferential interpretation given to these images,he decided in 1986 to enlarge them from 18 ×24 cm prints to 165 × 210 cm prints, to replaceindividually selected background colours withan evenly lit grey background and to concentrateon frontal poses, all of which added to the anon-ymising effect fostered by the inexpressivenessand seriality of the larger-than-life faces. The Por-traits series is not only paradigmatic of Ruff’sown work, but of several shifts in art photogra-phy’s display strategies and self-identificationthat took place in the 1980s: from small scale tolarge scale; from black and white to colour; andfrom photographic documentarism to conceptualart.

Figure 1. Thomas Ruff, jpeg ny02 (2004), C-Print, 269 × 364 cm (edition of 3).Source: Courtesy of the artist.

Hoelzl and Marie

80

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

In the 1990s, Ruff’s work took another signifi-cant turn: he no longer took pictures himself, butinstead used what we could call ‘already circulat-ing’ images from various sources and genres toconstitute his series. For his Sterne series (1990–1992), he purchased and enlarged negatives froman observatory. For Maschinen (2003), he pur-chased, enlarged and partly coloured glass nega-tives from a Dusseldorf machine manufacturer. Ina similar fashion to Gerhard Richter’s Atlas,started in 1964, his Zeitungsfotos (1990–1991)were taken from an archive of newspaper clippingsthat Ruff had collected over the years. Since Ruffcould not access an actual criminal photo archive,he used his own archive of portraits in his AnderePortrats series (1994–1995), which he superim-posed to create blurry images that resembledpolice sketches (Winzen 2001, 141–144).

In the late 1990s, the artist began using imagesfrom the web. In 1996, only two years after thelaunch of Netscape, Ruff began harvesting porno-graphic websites for his nudes series (2003).The series consists of small preview imagesthoroughly reworked in Photoshop by addingpixels and transformed into large-scale blurrycolour photographs of a veiled erotic appeal; in2002, he started his JPEGs series by organisingimages found on the web into different categories:‘catastrophes created by man’ (Figure 1); ‘cata-strophes created by nature’, ‘nature coming backor conquering man-made creations’ (Figure 2)(Lane 2009, 141). The series is complemented bya final category called ‘idylls’, which consists ofimages taken by Ruff himself on an artificialholiday island where he was vacationing and inthe Black Forest (Figures 3 and 4).2 The defaultsaving mode in small consumer digital camerasand the default mode of web images being JPEG,both these snapshot images and web-sourcedimages were, before any intervention by theartist, already submitted to the same compressionand storage method, the JPEG codec. To all theseimages the artist applied the same method of down-sizing, compression, interpolation and enlarge-ment, thus resulting in large-scale pixilatedtableaux that expose the new, digital ‘grammar ofphotography’ (Maggia 2006b), JPEG compression.The tableaux display, which we could call an aes-thetics of ‘hyperjpegness’, turns a compressionartefact into a cultural artefact.

While the nudes series addressed a veryspecific genre, pornography, the thematic scopeof the JPEGs is much larger. And while for thenudes series Ruff claims to have used blurring‘because the material [was] so ugly’ (Werneburg,taz Berlin, June 8, 2000), for the JPEGs he usesthe pixilation artefacts caused by JPEG com-pression to investigate the very aesthetics of theJPEG image. Both techniques, however, blurringand pixilation, are techniques used in the media(and in online services such as Google StreetView) to preserve anonymity. Unlike the nudes,the JPEGs do not feature any living humanbeings, and very rarely corpses, so that we arebrought to think that it is not the human being

Figure 2. Thomas Ruff, jpeg rl05 (2007), C-Print, 269.2×184.8cm (edition of 3).Source: Courtesy of the artist.

CODEC: on Thomas Ruff’s JPEGs

81

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

who becomes anonymous in the process, butrather the image itself. But what is an anonymousimage? We could say that it is not only an imagewithout an author, but a representational imagethat is non-representational, an image that refersto a specific place and time that at the same timeis floating in a spatiotemporal void which is the‘bare space’ of the photographic print. The anony-mous image no longer represents an absent object(as has been the meme for photography), but rep-resents the absence of an object.

Ruff’s method of maximum compressionresults in the appearance of normally undesiredblocks of 8 × 8 pixels, which represent the‘minimum coded units’ (MCUs) into which theimage is split during JPEG compression. Thehigher the compression, the more evident thevisual discontinuities between adjacent pixelblocks. One easily obtains the same effect usingthis method of resizing and applying high com-pression to a random digital photograph. Weresized a holiday photograph (Figure 5) to asmall file of 240 × 180 pixels and applied ‘zeroquality’ JPEG compression (Figure 6). To thenaked eye, the resulting image appears to behighly pixilated, even on a small print.

Figure 4. Thomas Ruff, jpeg en02 (2007), C-Print, 248.9 ×185.1 cm (edition of 3).Source: Courtesy of Gallery David Zwirner, New York.

Figure 3. Thomas Ruff, jpeg ri02 (2007), C-Print, 188 × 322 cm (edition of 3).Source: Courtesy of the artist.

Hoelzl and Marie

82

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

Figure 6. Remi Marie, beach (after Thomas Ruff) (2010), JPEG image, 240 × 180 pixel, 96 KB.Source: # Remi Marie 2010.

Figure 5. Remi Marie, beach (2010), JPEG image, 2048 × 1536 pixel, 1.4 MB.Source: # Remi Marie 2010.

CODEC: on Thomas Ruff’s JPEGs

83

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

By resizing and compressing web images andby ‘decompressing’ them as large-scale photo-graphic prints, Ruff brings the pixilation associ-ated with web images to the forefront of theimage. As photographic prints, the JPEGs rep-resent, via enlargement and media transferal, the‘abstract pixel structure’ of the JPEG image. Inan interview with Guy Lane, Ruff recalled:

Around 2002, I discovered that if you compressdigital files using JPEG compression, thesystem creates what I felt was a very interestingpixel structure. So I started an investigation,asking myself—How does it work? Wheredoes it come from? And then I decided tocreate whole images with this kind of abstractstructure (Lane 2009, 136).

In the same interview, Ruff related the method heused to give the JPEGs this particular quality: ‘Ihad to re-scale the files to a very small size andthen compress them as the worst possible qualityJPEGs’ (Lane 2009, 137).

Ruff’s work did not emerge in a void, however,as electronic artists had been exploring the aes-thetics of the digital artefact or glitch fordecades, starting with Nam June Paik’s manipu-lation of TV images in the 1960s and includingemerging graphic interface glitch artists such asJon Satrom. Nonetheless, Thomas Ruff was thefirst to bring the new mode of digital storage/dis-tribution of the photographic image (JPEG; web)and its new aesthetics (pixilation) into the fieldof art photography.

Here, we take Thomas Ruff’s aesthetics ofhyperjpegness as being paradigmatic of thecurrent state of digital visual culture as a whole,which is characterised by the dislocation of theimage from print to screen, from the archive tothe web and by the need for fast data transfer andaccess via enhanced network infrastructures(fibre optics, mobile broadband) and image infra-structures (compression). The sober title of theseries, characteristic of Ruff’s general titling prac-tice, describes both the subject matter (the JPEGimage as the new generic term for the photographicimage) and the artistic method (JPEG ‘hypercom-pression’). The pixilated images obtained through

this method also testify to a general shift in howphotographic images are stored and transmittedtoday: as digitally compressed files on the web.

The series eventually asks the question of thejpegness of JPEG images. Is JPEG compressionjust a technical way of storage or is it part of anew ontology of the image? Are the imagesfound on the web different from the images shotin the world or are they—as JPEG images—sub-stantially the same? Are compression and decom-pression protocols a substantial part of the imageor a technical envelope that can be added andremoved at will? Is the JPEG codec the new para-digm of the photographic?

3 JPEG and ‘the photographic’

JPEG stands for the Joint Photographic ExpertsGroup. Between 1986 and 1994, this technicalcommittee created the first international standardfor still image compression in order to ensureglobal compatibility. JPEG compression iscurrently the main mode of still image storage,transmission and display. It balances precisionand detail, which entails huge and slow files,against the need to save time and space via theelimination of redundant data and data that is notdetectable by the human eye.

The term ‘Joint’ refers to the link between theInternational Standardisation Organisation (ISO),which develops standards in a wide range offields from freight container dimensions to ISBNnumbers for books, and its specialised partnerorganisations, the International ElectrotechnicalCommission (IEC) and the International Telecom-munication Union (ITU), a United Nations agencythat develops international standards for infor-mation and communication technologies (ICTs).

The JPEG is a compression/decompressionprotocol, which is also known as a codec. Acomplex algorithm defines how a digital image iscompressed for storage and transmission andhow it is decompressed for display. The JPEGcodec was first published in 1992 as ‘ITU-TRecommendation T.81 (09/92)’, and in 1994 as‘ISO/IEC IS 10918-1:1994’.3 Although twentyyears have passed since its first publication, the

Hoelzl and Marie

84

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

JPEG-1, as it is called now, remains a standard partof most consumer applications.4

The gathering of the JPEG group is indicativeof the fact that the photographic image is a keyelement in the development of digital informationand communication technologies. The ‘photo-graphic’ in Joint Photographic Experts Groupstands for ‘continuous-tone still images’, whichare images with smooth transitions such as paint-ings, photographs or drawings, in contrast to ‘dis-crete tone images’ such as line drawings. For theJPEG group, the ‘photographic’ is no longer tiedto a specific recording and printing technology;instead, the term designates an array of digitalimages that can be compressed in the samemanner. Put differently, ‘photographic’ is a par-ticular aesthetic distribution (continuous-tone) ofpixels that can be correlated with each otherduring the compression process. For instance, ifone pixel of a photographed sky is blue, there isa high chance that the adjacent pixel will beclose to that blue; hence, the challenge of com-pression is to describe these spatial correlations(that translate physical continuities) in mathemat-ical terms.

Common file postfixes for JPEG compressedimages are ‘JPEG’ or ‘JPG’, while the images con-tained in these files are called ‘JPEGs’. The actualnames of the most common file formats are‘exchangeable image file format’ (‘exif’) and‘JPEG file interchange format’ (‘jfif’). Strictlyspeaking, JPEG stands for neither the image northe image file, but for the compression codec.Here, we will follow common usage and callJPEG compressed images JPEGs.

4 Codec/metacodec

The JPEG standard is implemented via a specificJPEG codec, i.e. a specific compression anddecompression algorithm geared towards continu-ous-tone still images. Let us briefly examine thefour main steps of the compression part of thealgorithm:(a) Down-sampling: encoding the RGB (red/

green/blue) information of the image usingthe YCbCr colour model, where Y is the

luma component and Cb and Cr are thechroma components. This allows for thesubsequent reduction of the spatial resol-ution of the Cb and Cr components, sincevariations in light are more perceptible tothe human eye than variations in colour.

(b) Block splitting: each Y, Cb and Cr channelof the image is split into blocks of 8 × 8pixels (Figure 7).

(c) Discrete cosine transform (DCT): each pixelblock of each component is converted to afrequency-domain representation; the fre-quency reflects the intensity and speed ofthe changes in luminosity and chrominancethat occur within a pixel block (Figure 8).

(d) Quantisation: each 8 × 8 frequency domainmatrix is divided by a new matrix called the‘quantisation matrix’, which is stored inmemory for the image decoding (Figures 9and 10). Dividing each 8 × 8 frequency-domain matrix by a constant and then round-ing it to the nearest integer greatly reducesthe amount of information in the high fre-quency components not detected by thehuman eye.

This rounding up that occurs during quantisa-tion is the only lossy operation in the entireprocess of JPEG compression. In the best case,‘lossy’ compression refers to a purely mathemat-ical loss, as the JPEG algorithm is gearedtowards what Buonori and Liberati (2008) havecalled ‘visual losslessness’. In other words, likethe 35 mm negative reduced image resolutionaccording to the standard display size of theimage, JPEG compression aims to balance datareduction with perceived image quality.

With high-level compression, the differencebetween two quantisation matrices applied to twoadjacent pixel blocks alters the continuity of theimage; the blocks appear as separate entities. By con-sciously using the highest level of image com-pression and the lowest image quality, i.e.resolution, Ruff takes advantage of this digital arte-fact. He forces the complex process of compressionto become visible as its normally undesirable sideeffect: pixilation. The technical default of the JPEGcodec—the apparition of visual discontinuities

CODEC: on Thomas Ruff’s JPEGs

85

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

between adjacent pixel blocks—is therefore trans-formed into a default aesthetic. This default aestheticis also the default (although mostly invisible) aes-thetic of the JPEG image in general, and thus of themajority of photographic images stored, distributed,5

displayed and viewed today—no longer in privatealbums, newspapers or on gallery walls, but on-screen and online.

But, in fact, the photographic dispositive—defined here as a chain of human–machine relationsthat includes all or some of the following sub-pro-cesses: recording/storage (or retrieval/storage);processing; distribution; display and reception/use6—has always included an element of com-pression and decompression, a ‘codec’, as it were.Applying this broad definition of codec to the photo-graphic image (and to perspectival representation ingeneral) helps to overcome the analogue/digitalfallacy and the separation of image and vision: wecontend that codec, or the ‘image as protocol’ is

not something unique to digital media, but hasbeen an aspect of image making and viewing eversince. With analogue photography, the recordingof the reflected light on a small negative is aspatial compression of information that isdecompressed via enlargement: photographicrecording (both analogue and digital) is the com-pression of a spatiotemporal extent to a two-dimensional image.

The mathematical principle of this technology,the linear perspective, reduces and flattens a three-dimensional object, locking it into a single(abstract) point of view and into a single (abstract)point in time. As Brunelleschi demonstrated inhis experiment (Manetti 1968, 44), with linearperspective, decompression is the task of theviewer who, when looking at this flat surface,restores its depth and temporality.7

In the JPEGs series, compression is at playnot only on an individual level (compressing the

Figure 7. Matrix of one block of 8 × 8 pixels, in Majid Rabbaniand Paul W. Jones (1991) Digital Image Compression Tech-niques, p. 116. Bellingham: SPIE Optical Engineering Press.Courtesy SPIE

Figure 9. Quantisation matrix, in Majid Rabbani and Paul W.Jones (1991) Digital Image Compression Techniques, p. 117.Bellingham: SPIE Optical Engineering Press. Courtesy SPIE

Figure 10. Quantised matrix, in Majid Rabbani and Paul W.Jones (1991) Digital Image Compression Techniques, p. 118.Bellingham: SPIE Optical Engineering Press. Courtesy SPIE

Figure 8. Matrix after DCT, in Majid Rabbani and Paul W.Jones (1991) Digital Image Compression Techniques, p. 117.Bellingham: SPIE Optical Engineering Press. Courtesy SPIE

Hoelzl and Marie

86

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

information of one image), but also on an overalllevel. All of the images pertaining to one ofRuff’s search categories, such as ‘nucleartesting’ or ‘icebergs’, in addition to all theimages that Ruff stored on his computer in thesecategories, are represented by only a handful ofimages bearing the same title code (e.g. ‘nt’ fornuclear testing and ‘ib’ for icebergs) and a numeri-cal code that does not exceed 05 (e.g. ‘nt02’ or‘ib01’ and ‘ib02’). The thousands of images thatRuff could have found on the web and the hun-dreds of images that Ruff did find and store onhis computer are compressed into one to fiverepresentative images. While the downsized andcompressed web images are only approximately100 kB, the images in the final print size of atleast 188 × 188 cm are expanded via interpolationto files of 1.3 GB. This means that each JPEG lit-erally comprises the data of approximately 10,000small web images.

Ruff’s entire method, including this two-stepprocess of image retrieval and selection, can infact be viewed as a metacodec. By metacodecwe mean that it is an artistic protocol that: (1) inthe etymological meaning of meta, comes afterand goes beyond the JPEG codec using bothhypercompression and what we could call hyper-decompression, i.e. the enlargement and printingof the final JPEG file as large-scale photographicprints (Figure 11); and (2) in the current meaningof meta, is about and on the JPEG codec, Ruff’sJPEGs speaking not only about their own codec

(just like an image’s metadata speak about itsdata), but on the JPEG codec in general.

5 The JPEG and the web

As previously mentioned, the majority of theimages in Ruff’s series were taken from the web,which suggests a certain relation between thesubject of the series, the JPEG codec and theweb. And, in fact, the JPEG codec’s publicationas a joint ISO/IEC standard in 1994 coincidedwith the launch of the first commercial webgraphic browser, Netscape Navigator, in 1994.This temporal coincidence is not surprising,since one of the main functions of the web is toshare information, first among high-energy physi-cists at CERN, then researchers worldwide, andnow an ever-growing number of the humanpopulation. The two inventions were in fact com-plimentary: without the web, the JPEG wouldhave been just another industrial norm; andwithout the JPEG, the web would not have devel-oped as it did. (We are, of course, aware that theweb not only hosts still imagery but also, andincreasingly so due to new internet technologiesthat speed up access and download,8 movingimagery; for these, separate compression stan-dards have been developed by the MovingImages Expert Group or MPEG.9)

However, the development of the web goesback to 1989, when Tim Berners-Lee, who atthat time worked for the Conseil Europeen pour

Figure 11. Installation views, Thomas Ruff. JPEGs, Gallery David Zwirner, New York, 2008.Source: Courtesy of Gallery David Zwirner.

CODEC: on Thomas Ruff’s JPEGs

87

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

la Recherche Nucleaire (CERN), wrote a propo-sal for ‘a large hypertext database with typedlinks’ that he first called ‘Mesh’ (Berners-Lee1989). By the end of 1990, Berners-Lee hadbuilt all the basic tools for the web as we knowit today: the HyperText Transfer Protocol; theHyperText Markup Language; a web browser(The World Wide Web); HyperText Transfer Pro-tocol server software; a web server (cern.ch); anda web page (http://info.cern.ch) that described theproject of merging ‘the techniques of informationretrieval and hypertext to make an easy butpowerful global information system’ (Berners-Lee 1991), finally naming it the ‘World WideWeb’ (www).10

While the web was first confined to the univer-sities and research labs that developed it, the intro-duction of the first graphical browser in 1993,called Mosaic and renamed Netscape Navigatorin 1994, marked a turning point towards commer-cialisation and popularisation. The first version ofMosaic lacked a return button;11 the first webbrowser was set up to only go forward, not back-ward (in this, the first version of the World WideWeb is similar to the early Google Street Viewversion, which did not have a built-in return func-tion). In a move that recalls the royalty-free divul-gation of Daguerre’s photographic process by theFrench Academie des Sciences in 1839, CERN,where the web was developed by Berners-Lee,declared in 1993 that the web protocol and codewere to be royalty-free.12

In the same line, the mission of the World WideWeb Consortium, which was also founded in 1993by Berners-Lee, was ‘to lead the World Wide Webto its full potential in developing protocols andguidelines that ensure the long-term growth ofthe web’,13 with the vision being ‘One Web’, anall-in-one communications tool that, throughmobile web standards, content adaptation, andmultimodal access, allows ‘any person to accessa specific information using any modalities onany devices from anywhere at any time’.14 Never-theless, this utopian vision did not prevent thelarge-scale commercialisation of web content,web access and user data that occurred in the late1990s and early 2000s with ‘free’ online services

such as Google Search and Google Maps, Skype,Facebook and Twitter, generating revenue via paidads, fee-based premium services and the exploita-tion of user data.

The World Wide Web Consortium defines theweb in communicative terms, as a ‘tool intendedto allow anyone, anywhere to share information’and in spatial terms, as ‘an information space’where every ‘item of interest’ is identified by aso-called ‘uniform resource identifier’, whichdefines the location or the name of a website.15

(The uniform resource locator (URL), theaddress of a website, is better known.) Unlikethe web, where the code for the transfer andwriting of hypertext is standardised, as are thebrowser and server software, the internet, com-monly defined as a global, publicly accessiblenetwork of computer networks, is not standar-dised—except for the TCP/IP protocol, whichbreaks data up into arbitrary packets that are thenrouted differently to their terminal where theyare reassembled for technical identifiers, IPaddresses and domain names.16

The word internet is in fact an abbreviation ofinternetworking, which was first used in the firstTCP/IP protocol developed by the NetworkWorking Group (1974) Request for Comments675; this means that before becoming a noun, theinternet was a verb—the description of an activityrather than of a structure. This resonates withAdrian Mackenzie’s discussion of wireless net-works in terms of the mobile relationality theyfoster, rather than in terms of the particular technicalstructure of wireless networks (Mackenzie 2011).

6 From archive to database

The temporal coincidence of the launch of Net-scape and the JPEG standard in 1994 initiated anew mode of storing and sharing digital imagesover the internet (a global network of computernetworks) and the web (a global repository oflinked data). In the two subsequent decades, archi-val material was dislocated in what could be calleda global ‘digitalisation fever’ from dusty andremote archives to remotely accessible databases.With this, the place and function of the archive

Hoelzl and Marie

88

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

have become the web and the database. But what isthe relation between the archive as a location ofdocuments and the web as an ‘informationspace’, and what is the relation between thearchive as an organised collection of documentsand the database as an organised collection ofdata?

Let us attempt to answer this question by goingback to a lecture that Jacques Derrida gave in 1994at the Freud Museum in London entitled Memory:The Question of Archives, which was later pub-lished as Archive Fever. Although not concernedwith the impact of the web on archival practices,Derrida’s definition of the archive as both territor-ial and performative is highly pertinent in account-ing for the issues at stake in the JPEGs series,whose source images are culled from the publicarchive of the web, organised into different cat-egories and stored in the artist’s personal archiveor database, and then finally transformed intoarchetypes via a complex process of downsizing,hypercompression, interpolation, enlargementand large-scale photographic printing. With this,the JPEGs constitute a new hybrid archive that isstored and accessed both physically (in themuseum or gallery) and digitally (on themuseum or gallery website and in both online oroffline databases).

In Archive Fever, Derrida (1994, 1–3) tracesthe meaning of the word ‘archive’ back to theGreek word arkhe, which means both commence-ment and commandment, linking it to the arkheion,the residence of the superior magistrates, thearchons, who command the city. The archons arethe documents’ guardians; they interpret the docu-ments and they speak the law. How does this relateto the idea and reality of the web as a global repo-sitory of linked data, of a ‘One Web’ that can beaccessed by anybody, anywhere, anytime andfrom any device? The fact is that, contrary to thehopes of media activists, as well as to the claimsof the World Wide Web Consortium, the web isnot a ‘Web for All’.17 While programmers can usethe web protocol and code royalty-free, the accessto web content is dependent on mostly fee-basedinternet access and often contingent on paymentor at least user registration. Like the Greek arkheion,

the web is well guarded. It is characterised by whatSaskia Sassen (1998, 177–179) calls ‘cyber-segmentation’, which means selective access toinfrastructure and content and the emergence of fire-wall-protected intranets.

From the Derridian perspective, cyber-seg-mentation could also mean the public provisionof indecipherable information. As Derrida (1994,2–3) explains, ‘the passage from private topublic is not always a passage from secret tonon-secret’). Documents containing sensitiveinformation, if they are given to the public at all,require interpretation by political, scientific or cul-tural experts. In ancient Greek society, the magis-trate replaced the priest; in a modern ‘informationsociety’ the magistrate has been replaced by theexpert. In a digital society, the expert is beinggradually replaced by software agents such asweb browsers and web crawlers and so-calleduser or personal agents that find and highlight‘important’ information on web pages or compilecustomised news reports, etc. It is the softwareagents (of which Google Search is certainly themost powerful) that decide, based on factorssuch as popularity, probability and personal prefer-ences, which information is ‘relevant’ to the user.

Ruff himself delegates agency to the GoogleSearch algorithm, which he uses to type insearch queries such as ‘icebergs’, ‘nucleartesting’ or ‘rocket launches’, and which preselectsthe images that he ‘finds’ on the web, even if partof his pictures result from serendipitous findswhen surfing from page to page. At the sametime, Ruff’s role here is, curiously and slightlyanachronistically, one of an image expert, a guar-dian and interpreter of images: from the publiclyaccessible portion of the World Wide Webarchive,18 which currently comprises billions ofimages,19 the artist creates a second and moreselective archive on his computer comprised ofbetween 5,000 and 10,000 images. From thisarchive, he selects images for a third archive, theactual JPEGs series, which comprises approxi-mately 160 images. The images are therefore sub-mitted to an accelerated process that archivalstudies refer to as ‘the theory of the three ages’(Perotin 1966). This theory refers to the three

CODEC: on Thomas Ruff’s JPEGs

89

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

functions and places that administrative docu-ments traverse during their lifespan:20 (1) thecurrent archives that are in constant use (inthe case of Ruff’s JPEGs series, the web); (2)the intermediate or semi-active archives whosecurrent use is limited, but which are temporarilyconserved (Ruff’s digital archive as a depositoryof potential JPEG images); and (3) the definitearchive (the JPEGs prints in the museum collec-tion).

In this process of selection and transformation,the images’ status is modified. Ruff ‘undresses’(Maggia 2006a, 143) the images of their photo-graphic parergon (author, time and place,caption). He strips them bare of nearly all infor-mation until they reveal the pixel grid of the digi-tally compressed photograph, and then restoresthem through interpolation and prints them aslarge-scale photographic C-Prints. In doing so,Ruff ‘redresses’ the images with new informationor a new meaning; they become what he refers toas ‘exemplary images’ (Lane 2009, 31), imagesthat represent ten, fifty or one hundred othersimilar images. As a result, they become arche-types in the very literal sense—from the Greekwords arkhe (principal, superior, first in time, orig-inal) and tupos (blow, impression, image, figure)as defined in the New Shorter Oxford English Dic-tionary (1993: 108 and 3441). They become, andthis is anachronous as well, primordial images,images that set the rules, that govern. As arche-types, the JPEGs are representational withoutbeing referential; they represent a type, not atoken. They realise what Latour (2010, 588) hasmistakenly called the ‘what of’ of pictorial rep-resentation in general: a tree; a missile; a high-rise; etc. As such, the JPEGs seem to be reducedto the mere gestalt or primordial form of whatthey depict, thus acting as an anchor or platformfor both a gestaltist perception and a vaguedeja vu, a palimpsest of visual memories, picturesseen in the media—and on the web.

The passage from archive to archetype alsoaffects the temporality of the photographic imagesthat constitute the JPEGs series: they are nolonger bound to the contradictory conception ofthe past as both irrevocably gone and technically

reproducible, and to the contradictory experienceof being ‘false on the level of perception, true onthe level of time’ (Barthes 1981, 63). Instead,they pertain to the atemporality of the arche-type—the first image—as generic or ‘exemplary’images that stand outside of time. They no longerrepresent a specific event (of the past), but a kindof event (past or future). For example, the imagesof 9/11 do not stand for the burning Twin Towersalone, but as Thomas Ruff himself put it in an inter-view, ‘for any building that was bombed during thelast 20 years—either by terrorist attacks or the USarmy or the air force’ (Lane 2009, 141).

The passage from archive to archetype is alsomarked by the archon’s interpretation. In Ruff’scase, the interpretation questions the very statusof the documents that are interpreted. As a result,the JPEGs contain two contradictory messages:(a) this is a just image, the just image of an eventor object that has been photographed; and (b)this is just an image, an ensemble of colouredsquares. This contradiction might be the veryground Ruff is operating on: the contradictionbetween the century-old dogma of photographicindexicality and the autonomy of the image as anensemble of picture elements, brought to the fore-ground by digital image processing, both on thelevel of the archive (selection of archetypicalimages and conversion of archive into database)and on the level of the individual image (JPEGcompression and pixilation artefacts). The JPEGsalso contains two contradictory messagesconcerning their value: while the method of hyper-compression contributes to the general devalua-tion of the image due to its increased availabilityand circulation on the web, the method of imageselection and elevation of the selected image tothe rank of ‘tableau’ contributes to its (financialand symbolic) revaluation.

7 On the algorithm

In his article, ‘The Algorithmic Turn’, Uricchio(2011) points out that from the first demonstrationof its operativity by Euklid in around 300 BC, thealgorithm remained a marginal form of calculationuntil the introduction of computers as algorithmic

Hoelzl and Marie

90

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

machines. Today, the algorithm is the underlyingprinciple of an entire range of digital applications,ranging from Wikipedia entries, interactive mapsand Amazon’s recommendation system, to auto-matic stock trading. Uricchio’s thesis is that thealgorithm announces a new way of organisingand using data via programmes, i.e. finite instruc-tion sequences that can be fed different values andyield different results. Whereas the modern Welt-bild is founded on precision, calculability and pre-dictability, resulting in a stable subject–objectrelation which finds its visual expression in a Car-tesian coordinate system, three-point-perspectiveand classical cartography, the algorithmic turnchallenges this representational order through the‘processual intermediation of subject-object rela-tionships’21(Uricchio 2011, 33), unstable pointsof view and collaborative authorship. This doesnot mean, as per Uricchio, that Heidegger’s Welt-bild, the conquest of the world as picture, is obso-lescent. But between calculating subject and cal-culated object, there now stands an algorithmiclayer. If we take his argument a step further, wecan say that the medium of the image is not thescreen, but in fact the operating algorithm, whichis responsible for the accumulation, organisingand displaying of data.

Uricchio refers to Heidegger’s (1977, 134)famous thesis, ‘The fundamental event of moder-nity is the conquest of the world as picture’, formu-lated in his essay ‘The Age of the World Picture’delivered in 1938. However, we can also readHeidegger’s essay differently, not in the sense ofpicture as a stable entity governing stablesubject/object relations, but as an operative toolthat allows one to act upon the world. Heideggerdefines ‘picture’ not in mimetic terms, but in thesense of ‘to set out before oneself and to setforth in relation to oneself’ (132). For Heidegger,the event of the world becoming picture coincidedwith the event of ‘man’s becoming subiectum inthe midst of what is’ (132), thus establishing aclose link between world picture and world view,defined as the ‘position of man in the midst ofall that is’ (134). We contend that this idea of theworld as picture set out before oneself in thesense of the object of one’s actions22 is not

limited to modernity, but that as a fundamentalhuman disposition of being in the world it can betraced back to Antiquity, e.g. to Ptolemy’s Geogra-phia being understood as the world’s first database.

Heidegger’s (1977) essay on the world pictureends on a cautionary note: man’s ‘unlimitedpowers for the calculation, planning, andmolding of all things’ (135), which are the basisfor the truthful representation of the world, reachout into the gigantic and thus turn into the ‘incal-culable’ (135), ‘an invisible shadow cast aroundall things’. Hence, ‘the modern world extendsitself out into a space withdrawn from represen-tation’ (136), which can be ‘known’ only by ‘crea-tive questioning’ and ‘genuine reflection’ (136).Thomas Ruff’s exploration of the technical arte-facts of JPEG compression can then be understoodas an artistic investigation of this ‘incalculability’and non-representability by means of ‘creativequestioning’. However, it is not the world and itsalgorismic representation that he questions butthe image and its algorithmic processing, whichis but a new way of calculating.

The algorithm used by Ruff (and by JPEGcompression in general) is a step-by-step pro-cedure used for calculation, data processing andautomated reasoning. Incidentally, this proceduresometimes fails in the case of unpredicted (andthus non-programmed) exceptions. One of themost spectacular failures of this kind took placein 1996 in the first Ariane 5 rocket, whichexploded only forty seconds after take-off,causing a loss of $7 billion in development costsand $500 million in material costs. The reasonfor this was a ‘software exception’ in the inertialreference system. In the report of the InquiryBoard published on 19 July 1996, one can readthe following remarkable explication:

The internal SRI (i.e. Systeme de ReferenceInertielle) software exception was causedduring execution of a data conversion from64-bit floating point to 16-bit signed integervalue. The floating point number which wasconverted had a value greater than whatcould be represented by a 16-bit signedinteger (Lions 1996).

CODEC: on Thomas Ruff’s JPEGs

91

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

The question of failure as being linked with thealgorithm is a very productive one: an algorithmhas to be tested, like a bomb, and here Ruff’simages of nuclear testing seem to come fullcircle. As we have seen with Ariane 5, a failurein an algorithm can cause catastrophes, infinitesi-mal as well as enormous: while the JPEGs seriesas a whole takes advantage of a failure of theJPEG algorithm that causes pixilation, the sub-series of the catastrophe depicts exemplaryimages of catastrophes. The JPEGs, as BennettSimpson puts it in his introductory essay to theJPEGs book, ‘depict breakdowns and are them-selves breakdowns of detail’ (2009). Conse-quently, the catastrophe is part of the image’sontogenesis; or to speak with Jacques Derrida(1993), the ruin is already there before theimage.23

However, the ‘breakdown of detail’ throughJPEG compression occurs only during quantisa-tion and is in most cases restricted to mathematicalloss, while on the visual level nothing appears tohave been lost. (The case of the JPEGs is different,since the goal here is in fact the very visualisationof the mathematical loss that is transformed by theartist into an aesthetic gain.)

But something is irrevocably lost with JPEGcompression, namely the very definition of thephotographic image: with the compression algor-ithm, which was standardised by the Joint Photo-graphic Expert Group, the photographic has beengeneralised as the bitmap or raster image, i.e.every digital image with the exception of thegraphic or vector image.24 With this, the new para-digm of the photographic is the algorithmic. Animage is no longer the product of a geometric pro-jection, but ‘a manifestation of algorithmicactivity’ (Marks 2010, 163).25 With JPEG com-pression, the geometrical ‘world projection’(Cavell 1979, 40) that informed the painterly orphotographic tableau gives way to the mathemat-ical abstraction of the numerical tableau. InFrench, tableau also means matrix. The image isno longer a tableau in the sense of a ‘stable archi-tectural order’ but in the sense of a digital matrix.The new architectural order of the image, then, isthe digital matrix of JPEG compression.26

8 Conclusion

Via a close reading of Thomas Ruff’s JPEGs seriesand the JPEG compression algorithm, we haveseen that the medium of the image is no longer aphotographic apparatus, and not even the digitalscreen where the majority of images are viewedtoday, but the compression and decompressionalgorithm that allows the image to be stored, trans-mitted and viewed. With this shift from geometricprojection to algorithmic processing, from imageas a product to image as a process, the transpar-ency claim is not obsolescent. We still view andunderstand the photographs we see on our digitalscreens as photographs, i.e. we look at them bymeans of looking through them at the objectsthey represent. But with digitalisation, the infra-structure of the image has changed, an infrastruc-ture which most of the time goes unnoticed andcomes to the fore only in the form of artefacts—errors and artistic creations.

This is where Ruff’s photographic work inter-venes—in creating opaque images, images thatcan no longer be looked through but that must belooked at. The JPEGs series does even more: itshows, in the form of discontinuous pixel blocks,that the new architectural order of the image isthe mathematical matrix used during image com-pression. With this, the meaning of the photo-graphic has radically changed—it no longerdenotes a specific mode of image creation butrather a specific mode of image processing. Withthe JPEG standard, the (still) image is no longerthe product of geometric projection, but the mani-festation of a compression/decompression algor-ithm.

Notes1 We borrow this term from Giampaolo Bianconi’s

article, ‘Gifability’ (2012), where he mentions thegrowing importance of animated and looped GIFfiles (low-resolution file format for web graphics)on Tumblr and other social media networks. TheseGIFs (of movies, TV shows and other mediaevents) are part of an internet-born, low-brow/activist culture of ‘memes’ (Goriunova 2011); theyexist, in a significantly higher resolution and high-brow context, as art (see Angelus Novus GIF by

Hoelzl and Marie

92

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

Garrett Rosenblum, 2012), but remain an essentiallypop cultural/internet phenomenon.

2 The categories are somewhat unclear: in an emailinterview with the authors (2 February to 4 April2011), Ruff explained that he classifies the imagesaccording to two main categories (catastrophes andidylls, which both have many shades) and to thespecific locations he searches for on the web.

3 See: http://www.itu.int/rec/T-REC-T.81-199209-I/en for the ITU-T Recommendation, and http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=18902 for the ISOstandard. The ISO standard PDF file can bepurchased for CHF 252, and the ITU-TRecommendation PDF can be purchased for onlyCHF 61. Updates are free. The JPEG homepagedirectly refers to the ITU-T download. See: http://jpeg.org/jpeg/index.html.

4 JPEG 2000, which was first launched in 2001, offersthe choice between a lossless alternative (1:2) andstandard lossy compression (1:10 for greyscaleimages; 1:20 for colour images). JPEG 2000images are smaller and of higher quality thanJPEG images, and do not display pixilationartefacts owing to a different compression method(which can, however, cause blurring artefacts).Moreover, the JPEG 2000 standard allows thestorage of images with 48-bit colour depth (similarto TIFF images), the inclusion of metadata such asgeographical coordinates within the image file, andthe specification of so-called regions of interest(ROI) that can be saved with a higher resolution.The wavelet transform method progressivelyencodes the image, yielding a multi-resolutionimage which creates the separate saving of anoriginal TIFF file, a medium resolution JPEG foronline display and a low-resolution thumbnailsuperfluous (Buonori and Liberati 2008).

In their study on JPEG 2000 file robustness, PaoloBuonora and Franco Liberati argue even for the lossyJPEG 2000 as a new archival format for high qualityimages, since it is not mathematical losslessness thatcounts but visual losslessness. But since JPEG 2000requires more computational power, is not entirelylicence free, and requires expensive hardware, mostcommercial manufacturers and software developersstill use the old format. JPEG 2000, however, isused in particular contexts such as medical imaging,digital cinema and long-time archiving (Buonori andLiberati 2008). Somewhat ironically in the case ofThomas Ruff, who started his career with hisPortraits in the style of passport photographs, the

JPEG 2000 standard is also used for the newGerman passports (Friedrich 2006).

5 The particular thing about web storage is that, bydefault, web storage means web distribution. Thedefault mode of MyMaps on Google, as well asthe default mode of Twitter messages, is ‘public’.However, these services all offer private modes.With cloud computing gaining ground, the weband its services might in fact split into two parallelworlds: one of linked public data and one ofprivate data accessible only by individual usersand invited group members. However, we shouldnot forget that ‘private data’ are always shareddata, since they are at least shared with the ISP andthe application one uses.

6 With this definition, we are acknowledging bothMichel Foucault’s (1980) expanded definition ofthe dispositive as the technical and social apparatusof power relations and Jean-Louis Baudry’s (1975)definition of the cinematic dispositive comprisingthe projector, the screen and the viewer. Ouremphasis is on the succession of the different stepsin the ‘production’ of the photographic image,from recording to reception/use. See also NicolasBourriaud’s (2002) reference to ‘post-production’,which includes the active reception or use ofartefacts.

7 Whether the competence of correct decompression,i.e. the correct interpretation of the perspectivalimage, is acquired or is a so-called‘anthropological constant’ is still subject to debate.

8 With streaming technology, video imagery (inparticular, TV broadcasts and data-intensivemovies) can be watched at the same time as it isbeing downloaded, thereby increasing the user’ssensation of real-time interaction with onlinecontent.

9 The Moving Pictures Experts group published itsfirst standard in 1993. MPEG-4, or ISO/IEC14496/ITU-T H.264, was developed between1998 and 2001. Like the JPEG, the MPEG is ajoint venture of ISO (International StandardisationOrganisation, IEC (International ElectrotechnicsCommission) and ITU (InternationalTelecommunications Union). Unlike the JPEGstandard, the MPEG standardises only thebitstream and the decoder, but leaves the encoderto commercial developers. See the MPEGhomepage, http://mpeg.chiariglione.org/standards/mpeg-4/mpeg-4.htm.

10 This name choice privileged and anticipated theglobal expansion of the web. Other options thatwere turned down by CERN were The Information

CODEC: on Thomas Ruff’s JPEGs

93

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

Mine (which could be read as TIM, the developer’sfirst name) and the Mine of Information (whichcould be read as MOI in French). Both highlight theaim of the first web browser as a data-miningmachine.

11 http://en.wikipedia.org/wiki/History_of_the_World_Wide_Web.

12 The original document can be viewed here: http://tenyears-www.web.cern.ch/tenyears-www/Declaration/Page1.html.

13 http://www.w3.org/Consortium/mission.html.14 http://www.w3.org/standards/webofdevices/

multimodal.15 http://www.w3.org/Help/#webinternet.16 For a brief technical survey of the different levels of

the internet from the physical level to the level ofuser interaction, see the first chapter of AlexanderGalloway and Eugene Thacker’s, The Exploit. ATheory of Networks (2007).

17 http://www.w3.org/Consortium/mission.html.18 As is well known, the ranking algorithm of Google

Search ties relevance to popularity; popular resultswill be shown first, which is a self-assertiveprocess: since documents that deal with rare issues,or documents that are poorly hyperlinked, willdisplay only on the far end of the Google Searchresult ranking, they will not be accessed, whichmeans that they are further sliding back into thedusty corners of the web archive. And then thereare pages that are not even listed among theGoogle Search results, or that are not even listedon the Google Website index. Can those pages stillbe considered to be part of the web archive,following the definition of the archive as anorganised collection of documents?

19 The photo-sharing website Flickr owned 5 billionimages as of September 2010. See http://blog.flickr.net/en/2010/09/19/5000000000/, which announcedthe 5 billionth uploaded image.

20 The theory of the three ages was formulated in 1948by a working group from the Hoover Commission(officially named the Commission on theOrganisation of the Government). This theorygreatly influenced archival practices in the secondhalf of the twentieth century. For more informationon the theory of the three ages, see Perotin (1966),and on its criticism see Caya (2004).

21 Uricchio elaborates (2011, 33): ‘A process of neitherthe subject’s nor the object’s making determinespoint of access, sights seen, connections made,experience gained. This is done, in a manner ofspeaking, in collaboration with both the subjectand the object.’

22 In his essay ‘Eye and Mind’, Maurice Merleau-Ponty(1964, 160) sharply criticises the scientific or‘operative worldview’, wherein the world isunderstood as ‘the object X of our operations’,which stands in sharp contrast to the bodily andholistic experience of the world that Merleau-Pontyadvocates. With Heidegger, we contend thatoperativity is not merely an abstraction andoperationalisation of the world, but in fact adisposition that allows humans to be in and withthe world.

23 Simpson’s article is entitled ‘Ruins: Thomas RUFF’sJPEGs’, but does not directly reference Derrida.

24 See Sean Cubitt’s keynote lecture, ‘Time to Live’, heldat ISEA 2011, Istanbul. For Cubitt, the relationbetween the raster and the vector image is that of apolitical struggle of emancipation: ‘The job at handis to investigate which emergent principle orprinciples stand opposed to the hegemony of the gridand its spatialisation of everything’ (Cubitt 2011,14–15).

25 This argument is made in the context of Marks’compelling comparison between Islamic art andnew media art. ‘In computer media as in Islamicart’, Marks argues, ‘image is a manifestation ofalgorithmic activity. However, that algorithm mayremain inactive, and the image may remain latent’(Marks 2010, 163). In the case of JPEGcompression, we can argue that every raw imagefile is a latent JPEG compressed file and everyJPEG compressed file a latent decompressed file.

26 We refer here to Francois Chevrier’s (2003) notion ofthe tableau as a stabile architectural composition thathas been shattered through the advent of the sketch,photographic technique, and later collage, etc. As perChevrier, late 1980s photography restores the tableauformat, no longer in the intention of it containing theentire world, but as a fictional frame. In particular,Chevrier’s points of reference are the Becherstudents, Thomas Ruff, Andreas Gursky andCandida Hoefer. For Chevrier, the ‘tableau-form’means a certain ‘return to order’, albeit a lucid one;here, we argue for a change of order: from thetableau as a graphic composition to the tableau as amathematical matrix used for image compression.

ReferencesBarthes, Roland. 1981. Camera lucida: Reflections on

Photography. New York: Hill and Wang. (First pub-lished 1980 as La Chambre claire: Note sur la photo-graphie. Paris: Gallimard.).

Hoelzl and Marie

94

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

Baudry, J. L. 1975. “Le dispositive.” Communications23: 56–72.

Berners-Lee, Tim. 1989. “Information Management: AProposal.” Accessed February 22, 2013. http://www.w3.org/History/1989/proposal.html.

Berners-Lee, Tim. 1991. “WorldWideWeb – Summary.”Accessed February 2, 2013. http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Summary.html.

Bianconi, Giampaolo. 2012. “GIFABILITY.” Rhizo-me.org. Accessed February 2013. http://rhizome.org/editorial/2012/nov/20/gifability/.

Bourriaud, Nicolas. 2002. Postproduction. Culture asScreenplay: How Art Reprograms the World.New York: Sternberg.

Buonora, Paolo, and Franco Liberati. 2008. “A Formatfor Digital Preservation of Images: A Study onJPEG 2000 File Robustness.” D-Lib Magazine 14(7\8). Accessed February 2013. http://www.dlib.org/dlib/july08/buonora/07buonora.html.

Cavell, S. 1979. The World Viewed: Reflections on theOntology of Film. Cambridge, MA: Harvard Univer-sity Press.

Caya, M. 2004. “La theorie des trois ages en archivis-tique. En avons-nous toujours besoin?” Conferencesde l’Ecole des chartes. Accessed June 2011. http://elec.enc.sorbonne.fr/conferences/caya.

Chevrier, F. 2003. “The Adventures of the Picture Form inthe History of Photography (1989).” In The LastPicture Show: Artists Using Photography,1960–1982, edited by D. Fogle, 113–128. Exhibitioncatalogue. Minneapolis, MN: Walker Art Center. (Firstpublished, 1989, as “Les aventures de la forme tableaudans l’histoire de la photographie.” In Photo- Kunst:Arbeiten aus 150 Jahren, edited by F. Chevrier. Exhi-bition catalogue. Ostfildern/Ruit: Hatje Cantz).

Cubitt, J. 2011. “Time to Live.” In ISEA 2011 IstanbulConference Proceedings. Leonardo ElectronicAlmanac, 8–15. Accessed February 2013. http://www.leoalmanac.org/wp-content/uploads/2012/04/ISEA_proceedings-sean-cubitt.pdf.

Derrida, Jacques. 1993. Memoirs of the Blind: The Self-Portrait and Other Ruins. Chicago, IL: University ofChicago Press.

Derrida, Jacques. 1994. Archive Fever. A FreudianImpression. Chicago, IL: University of Chicago Press.

Foucault, Michel. 1980. 1977. “The Confession of theFlesh.” In Power/Knowledge Selected Interviews

and Other Writings 1972–1977, edited by C.Gordon, 194–228. New York: Pantheon Books.

Friedrich, E. 2006. “The Introduction of ElectronicPassports in Germany.” Paper given at the 2ndSymposium on ICAO-Standard MRTDs, Biometricsand Security. 6–7 September 2006, Montreal.Accessed 9 August 2013. http://legacy.icao.int/icao/en/atb/MRTDsymposium/MRTD_06/Presentations/Friedrich.pdf.

Galloway, Alexander, and Eugene Thacker. 2007. TheExploit. A Theory of Networks. Minneapolis, MN:University of Minnesota Press.

Goriunova, Olga. 2011. Art Platforms and Cultural Pro-duction on the Internet. New York/Oxford:Routledge.

Heidegger, Martin. 1977 [1938]. “The Age of the WorldPicture.” In The Question Concerning Technology andOther Essays, 115–154. New York: Harper & Row.

International Standardisation Organisation. 1994.“ISO/IEC IS 10918-1:1994”. Accessed June 2011.http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber¼18902.

International Telecommunications Union. 1992. “ITU-TRecommendation T.81 (09/92)”. Accessed June2011. http://www.itu.int/rec/T-REC-T.81-199209-I/en.

Lane, G. 2009. “Thomas Ruff: Space Explorer.” ArtWorld 12: 136–143.

Latour, B., E. Camacho-Hubner and V. Novembre.2010. “Entering a risky territory: space in the ageof digital navigation.” Environment and PlanningD: Society and Space 18 (4): 581–599.

Lions, J. L. 1996. “ARIANE 5. Flight 501 Failure.Report by the Inquiry Board.” Paris, 19 July1996.” Accessed February 2013. http://www.ima.umn.edu/~arnold/disasters/ariane5rep.html.

Mackenzie, A. 2011. Wirelessness. Radical Empiricismin Network Cultures. Cambridge, MA: MIT Press.

Maggia, F. 2006a. “Conversation with Thomas Ruff,April 2006.” In Thomas Ruff: The Grammar of Pho-tography, edited by F. Maggia, 142–143. Exhibitioncatalogue. Milan: Nepente.

Maggia, F. 2006b. “The Grammar of Photography.” InThomas Ruff: The Grammar of Photography,edited by F. Maggia, 139–142. Exhibition catalo-gue. Milan: Nepente.

Manetti, Antonio. 1968. The Life of Brunelleschi. Intro-duced and annotated by Howard Saalman. Philadel-phia, PA: Penn State Press.

CODEC: on Thomas Ruff’s JPEGs

95

Dig

italC

rea

tivity,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4

Marks, L. 2010. Enfoldment and Infinity. An IslamicGenealogy of New Media Art. Cambridge, MA:MIT Press.

Merleau-Ponty, Maurice. 1964. “Eye and Mind.” In ThePrimacy of Perception and Other Essays on Phe-nomenological Psychology, the Philosophy of Art,History and Politics, edited by James M. Edie,159–190. Evanston, IL: Northwestern UniversityPress.

Network Working Group. 1974. “Specification of Inter-net Transmission Control Program.” Request forComments: 675. Accessed February 2013. http://tools.ietf.org/rfc/rfc675.txt.

Perotin, Y. 1966. “Administration and the ‘Three Ages’of Archives.” The American Archivist 29 (3):363–369.

Rabbani, Majid and Paul W. Jones. 1991. Digital ImageCompression Techniques. Bellingham, WA: SPIEOptical Engineering Press.

Ruff, T. 2003. nudes. With an essay by Michel Houlle-becq. New York: Harry N. Abrams.

Ruff, T. 2009. JPEGs. With an essay by BennettSimpson. New York: Aperture.

Sassen, S. 1998. “Electronic Space and Power.” In Glo-balization and Its Discontents: Essays on the NewMobility of People and Money, edited by S. Sassen,177–194. New York: The New Press.

Simpson, B. 2009. “Ruins: Thomas Ruff’s JPEGs.” InJPEGs, edited by T. Ruff. New York: Aperture.

Uricchio, W. 2011. “The Algorithmic Turn: Photo-synth, Augmented Reality and the Changing Impli-cations of the Image.” Visual Studies 26 (1):14–24.

Winzen, M. 2001. “A Credible Invention of Reality.Thomas Ruff’s Precise Reproductions of Our Fanta-sies of Reality.” In Thomas Ruff. 1979 to the Present,edited by M. Winzen, 131–159. Exhibition catalo-gue. Cologne: Buchhandlung Walter Konig.

Ingrid Hoelzl is an Assistant Professor at theSchool of Creative Media at City University ofHong Kong. Her current research interests are in

the fields of ICTs, wireless and locative media,urban navigation, digital image and mediatheory, aesthetics, and digital urban living. She isthe author of a book on the theory of photographicself-portraiture, Der Autoportratistische Pakt(Wilhelm Fink, Munich, 2008), the guest editorof a History of Photography special issue entitledPhotography and Movement (volume 35, issue 1,February 2011) and the co-author (with FriedrichTietjen) of the anthology Images in Movement(LUCA, Brussels, 2012). Her articles on contem-porary photography, photographic self-portraitureand the new temporality and place of the imagein digital culture have been published in numerousbooks and exhibition catalogues, and in peer-reviewed journals such as Photographies,History of Photography, Intermedialites and Leo-nardo/Transactions.

Remi Marie is a French writer and performertrained first in physics and statistics, then in arthistory and aesthetics. Having worked as a setdesigner and assistant for cinema and dance pro-ductions, and as a painter and media artist, hestarted writing experimental literature in 2001.His autofictional texts and essays on art, politics,and globalisation have been published in numerousFrench, Austrian and Canadian journals includingLa Revue Litteraire, Le Quartanier and Nioques.He has performed (in collaboration with musicians)at venues such as the 2011 Act’Oral festival in Mar-seille (with Erik Minkinnen) and Camera Obscurain Montreal (with Vergil Sharkya). His first book,Je, was published by Leo Scheer in 2010. He is cur-rently working on two literary book projects:Thomas Neffe, an experimental translation ofThomas Bernhard’s Wittgenstein’s Neffe [Wittgen-stein’s Nephew], and Bien parti, an autofictionalaccount of the desires, politics and economics ofpostmodern globalised city nomadism.

Hoelzl and Marie

96

Dig

italC

rea

tivity

,V

ol.

25

,N

o.

1

Dow

nloa

ded

by [

T&

F In

tern

al U

sers

], [

Car

ol A

lexa

nder

] at

05:

59 1

0 M

arch

201

4