‘Generating a new sense of place in the age of the metaview’, Journal of Australian Studies, Volume 35, Issue 4, 2011. With James Steele

A new sense of place in the age of the metaview

In the period since its invention, photography has become one of the main means through which we create and maintain our sense of place. For a long time photography has grounded our human evocations of a site’s cultural, historical or natural significance to its actual physical location. As Joan M Schwartz and James R Ryan put it, ‘Photographs shape our perceptions of place … photographic practices — from tourist photography to domestic photography — play a central role in constituting and sustaining both individual and collective notions of landscape and identity. Moreover, photography has long played a central role in giving such social imagery solid purchase as part of the ‘real’.’[1] For many of us it is our prior understanding of a site’s significance through seeing photographs of it that we feel to be so strongly confirmed by our own eventual experience of its actual material reality. For instance we understand the significance of the Tasmanian wilderness through all the photographs of it by Peter Dombrovskis and Olegas Truchanas. And we process our experience of the sheer physical presence of Uluru through all the thousands of photographs of it we have seen.

If photographs have been turning spaces into places for such a long time, what can be said about that process now after two revolutions, the first being the digitisation of the film-based imaging in the 1990s, and the second being the explosion of online uploading, archiving, access and distribution of photographs on the web in the 2000s. How have these revolutions affected the framing, sharing and experiencing of places?

THE FRAME

With the invention of photography the convention of the wooden frame around a painting, which had been developing in the West for a thousand years, became technologised and internalised into the camera itself, first into the glass-plate holder, and then into the viewfinder which punched a rectangular ‘picture’ out of the circular image thrown by the lens. Before photography frames were wooden artefacts that separated the art from the architecture in order to, in Immanuel Kant’s words, ‘stimulate representation by their charm, as they excite and sustain the attention directed to the object itself’.[2] After photography, however, framing became a more instrumental act that concentrated and directed the eye of the viewer. In the case of landscape photography this framing act incorporated the conventions already developed in landscape painting. The succeeding planes of mountains and hills, the compositional vectors of valley and rill, led the viewer’s eye into the virtual space of the picture. Many subsequent critiques of European imperialism have also seen this act of scopic framing and penetration as emblematic of the commandeering and domination of the land itself.[3]

In the case of stereo views the narrativization of the enframed space almost became violent as the viewer seemed to be virtually jolted out of their chair and phenomenologically drawn into the 3D illusion of the picture. As the essayist Oliver Wendell Holmes enthused to his readers in 1859: ‘The first effect of looking at a good photograph through the stereoscope is a surprise such as no painting has ever produced. The mind feels its way into the very depths of the picture. The scraggy branches of a tree in the foreground run out as if they would scratch our eyes out.’[4]

Photography brought to the forefront the frame’s capacity for not only gathering up, concentrating and organising the viewer’s attention, but also for — again almost violently — cutting what was inside the frame off from the reality that continued to extend beyond it. By the mid twentieth century to compose was also to crop. As John Szarkowski remarked in 1966 ‘The Photographer looked at the world as though it was a scroll painting, unrolled from hand to hand, exhibiting an infinite number of croppings — of compositions — as the frame moved onward’.[5] Szarkowski’s conventional art-historical analogy is certainly well chosen as a means of acknowledging photography’s close historical relationship to both Western and Oriental painting traditions, but it misses the full agility of the camera as the photographer swooped the camera up and down and zoomed in and out. The slicing edge of the photograph added a graphic intensity to the spatial drama of the view, boxing spaces up with shapes that were amputated and truncated by the frame. Frames were particularly good at giving monuments and landmarks a graphic iconicity. For instance the conventional postcard framing of Uluru sitting on the horizon has a flag-like recognisability.

METADATA

Photography had always been a medium of quantity. From the beginning photographs were packaged into albums to be circulated, or reproduced in books to be published. The digital revolution and the rise of the internet exponentially accelerated that old order. Almost five billion images have been uploaded to the photo-sharing site Flickr since it began in 2004[6]. Over two billion photos go up onto Facebook each month.[7] The central object of the medium, the photograph as artefact — be it daguerreotype, stereograph, print or book – is disappearing under the weight of the simplicity, accessibility and reach of the new, digital form. New possibilities opened up by the online digital environment have changed the way images communicate and audiences are moving to new virtual environments to experience images. New methods of finding, sorting and viewing images are being developed. Ease of access to these images is creating new technological opportunities to assemble and present transient collections of related images. Such collections can be easily shared so others can experience them, or in their turn actively reorder, expand or limit the images, how they see them, and what meaning they derive from them.

Although algorithmic recognition software, such as face recognition software, is beginning to be developed to search for images by directly matching the features in the images themselves (colour, faces, shapes and so on), the most common way of locating images on the internet is still through simple text searches. [8] To find an image through a text search the text a user seeks to match with a photograph must have somehow been previously associated with the image in a way that makes sense to a search engine. Typically, the title of the photograph, its description or text surrounding an image embedded in a web page is used to make the match with search terms. This associated text is metadata: data (text descriptions) about data (the image) that makes access to the image possible through text-based searches. There are other sorts of metadata as well, for example, librarians catalogue photographs with a range of descriptors, while whenever a picture is taken digital cameras automatically save data about such things as the shutter speed, aperture, camera type, lens type, and the date and time the exposure was made.

Search engines can use this extended metadata to aid user searches. Specialist sites like the National Library of Australia’s Trove allow users to search catalogue data to find items that interest them in the collections not only of the National Library itself, but also other institutions across Australia and even beyond: books, newspapers, journals, maps, music, oral history, paintings, photographs and more besides.[9] Items are indexed by Trove to help searchers find items on Australia and Australians. Trove is a portal or metasearch engine, providing access to items using data provided by a range of sources, many of which are external to the National Library and out of its control.

Flickr uses a similar approach in its Commons Project.[10] Flickr Commons exposes the photography collections of a number of large institutions, including the Australian War Memorial, State Library of New South Wales and Sydney’s Powerhouse Museum. In a process known as crowdsourcing, Flickr encourages its community of users to add their own tags and comments about the photographs they see in the Commons. The tags and comments become part of the public data associated with the photographs, discoverable through searches by other users.

Another type of metadata that can be attached to items like photographs is the location where the photograph was created. With digital photographs, this information can be embedded electronically in the image file itself, along with other standard metadata like the date and time the photograph was taken, any keywords associated with the image, copyright information and so on. As the digital photograph is shared across the internet, so too is the embedded metadata. Digital photographs with location data embedded in them are said to be ‘geotagged’.

When the Panoramio layer is turned on in Google Earth, small icons, or the images themselves, appear on the surface where geotagged photographs were taken. Clicking the icons or images displays the photograph. Icons also appear for other geotagged photographs that were taken nearby, and can be similarly displayed.

[INSERT PHOTO1 HERE]

Panoramio layer on Google Earth showing the presence of geotagged images along the Champs-Élysée

As more and more photographs have location metadata attached, it becomes easier to find different views of a place, or even the same view but taken at different times. Currently about eight million geotagged photographs are uploaded to Flickr every month.[11] These pictures can be searched and reconfigured to create other pictures, for instance in 2009 over 35 million geocoded photographs were harvested from Flickr to create a ‘heat map’ of the world showing where the photographs had been taken[12].

[INSERT PHOTO 2 HERE]

World Heat Map: http://www.cs.cornell.edu/~crandall/photomap/, used with permission

METAVIEWS

What is emerging is the possibility of software applications within online environments creating an enhanced sense of a place by using the increasing numbers of photographs available of a particular location. A single photograph is one moment in time, while the more photographs there are of a place, the greater the possibility of capturing the events, the structures, and the beauty of the place that provides meaning and defines it as something worth experiencing. An even richer sense of place can be further revealed through the comments added by the user community.

Before digital photography, snapshot photographs circulated in different spheres to official photographs, and historic photographs were kept in different collections to contemporary photographs, but now they are all just a hyperlink from each other. Different images of the same place taken at different times can help us build up a more complete ‘picture’ of the activities and events that happened there, giving us a new sense of the place that would be impossible to extract from any one image. Traditionally, different photos were looked at individually, on the wall of a gallery, in the pages of an album or book, within a pile of holiday prints back from the chemist, or at a slideshow on a Saturday night. Just as the Renaissance picture frame was remediated into the camera, these old forms of viewing have been remediated into the internet, with its document, page, slide-show and album metaphors. But computers are also capable of previously impossible forms of display: using software, similar images can be combined, stitched together in panoramas, used to create tone-mapped images, or laid out in three-dimensional space so that viewers can navigate through them almost at random.

Through the use of these online technologies viewers now have the opportunity to experience extended views of landscapes that grow beyond the rectangular frames of cameras. These views can be called ‘metaviews’ for several reasons. Firstly because they are not framed by a rectangular boundary cropping a composed scene out of an infinitely extensive reality, rather they are ‘enframed’ by search terms applied to the metadata attached to each picture in an online archive. This sifting of the metadata may gather together pictures that share a common locality established by proximate GPS co-ordinates for instance, or a similar compass direction, or a similar time of the year, or the same key word in the caption. The viewer’s eye composes a sense of a view by travelling through a sequence of photographs that erase the frame. Just as the eye was led through traditional landscapes, or the viewer apparently entered the very depths of the stereographic view, in the metaview the user navigates through a contiguous collection of images.

The experience of the view they get is therefore a second order view, a metaview in a second sense because it is constructed from first order components — photographs. It is not, however, a slide show or a cinematic or paginated sequence because it is not a deliberately authored enunciation, rather it is a technological enframing which in and of itself ‘excites and sustains’ the user’s attention which is directed to the metaview. These enframed views are ‘meta’ in a third sense because they become a virtual place within the space of the web, a place that, often, users can navigate themselves. They are another ‘site’ on top of the website the viewer is using. As William J Mitchell noted, increasingly ‘the places we frequent have IP addresses as well as geographic coordinates.’ [13]

Computer software currently exists where individual photographs can be stitched together to extend the frame by creating panoramas, or gigapixel images[14], where the viewer can pull back to see the whole scene or zoom in to any part for fine detail. High Dynamic Range (HDR) images are created using computer software by combining several photographs of the same scene taken with difference exposures, to bring out details in the shadows and the highlighted areas that normally would not be able to be distinguished with film or digital photography. QuickTime VR images[15] are spherical images that can be viewed in a web browser. They can show a full 360 degree view of an object or space, allowing the viewer to choose the perspective from which they want to see the scene. Google’s free SketchUp software[16] is a 3D modelling application that allows users to skin their 3D models with actual photographic images, providing the viewer with more angles to view the subject of the model than would normally be possible using individual images.[17] Microsoft’s Photosynth[18] and AutoDesk’s Photo Scene[19] place photographs taken of the same subject or scene from different perspectives in a three-dimensional space and provide the viewer with tools to navigate through the scene on the computer screen.

Currently, most metaviews are created from the collection of a single photographer, although there are some examples where online communities contribute images of a particular place or event, and the metapicture is created from the contributions of many people[20].  Unlike the traditional ‘decisive’ moment, these metaviews have the potential to span time, from microseconds to more than a century, giving viewers insights, for example, into the impact of human activity on the environment; insights that would be harder or impossible to achieve with individual photographs. Different points of view may reveal information hidden in the single framed photograph. In his article Walls of Light – Immaterial Architecture, Scott McQuire refers to ‘cumulative knowledge established by the series or set’[21]. Collections of images related by time and place add a dynamic dimension to the individual photograph, transcending the original frame and allowing us to peer outside it into space the original photographer was unable or unwilling to let us see, giving us a wider perspective and greater insights than would otherwise be possible. Multiple perspectives on the same place or event may provide viewers greater opportunities to be witnesses rather than consumers of packaged stories, allowing them the sense that they are informing themselves and making up their own minds.

A SENSE OF PLACE

If individual photographs of somewhere give us evidence of what happened or exists there, then many photographs, say of different events at that location or of the location through time, might provide more of the resonance that turns space into place. Previously, photographers have always sought to encapsulate the significance of a view in a single frame. The paradigmatic example in Australia is Peter Dombrovskis’ picture of Rock Island Bend on the Franklin River, reproduced in full-page colour in Australia’s newspapers on election eve 1983 with the slogan ‘Could you vote for a party that would destroy this?’, it was arguably responsible for  electing Bob Hawke’ Labor government.[22]  Other significant parts of Australia, such as Uluru, have become widely experienced through their amenability to iconic, graphic iconicity. The metaview is now available to be used to contribute to the material experience of more Australian places, perhaps those not so immediately amenable to spectacular encapsulation.

The first signs of this movement are beginning to appear. For example Picture Australia is the National Library of Australia’s service that allows users to search the pictorial collections of many of Australia’s collecting institutions.[23]  Individuals can also contribute photographs to the Picture Australia pool using Flickr: simply by joining a Picture Australia Flickr group, tagging their photos and contributing them to the group, those user-contributed photographs on Flickr with metadata matching a user’s search criteria will be returned to other users searching Picture Australia. The National Library also uses the Flickr groups to identify contemporary images that may be added to its own collections. With the sheer deluge of images coming available on the internet, the utilisation of Flickr contributors’ own efforts to tag and nominate photographs that fit the Picture Australia criteria gives the National Library a mechanism to cope with the flood.

A recent project with the Kosciuszko Huts Association (KHA) is an example of how the National Library has been able to harness the power of the internet to augment its limited resources for accessioning images. The National Library had previously accepted over 2 000 photographs, negatives, and slides contributed by KHA members, but the Library was not in a position to accession further photographs. Keen that all these additional images not be lost to the community, the Library suggested to the KHA that its members upload the images to Flickr and contribute them to the Picture Australia ‘Ourtown’ Flickr group. While the number of photographs actually shared by the KHA on the group remains relatively small, projects such as this nonetheless indicate the potential for groups with their own idiosyncratic histories and private collections to form larger searchable archives.

On 23 July, 2010, at Whites River Hut in the High Country a skier threw spirits into a pot-bellied stove to light it, not knowing that there were still glowing embers there. The resulting fire quickly took hold inside the hut, but a party of passing skiers extinguished it before the hut was completely destroyed[24]. The KHA sent out a call to its members to contribute photographs of the hut before the fire to its photography website so that a 3D model of the hut, outside and in, could be developed using Google’s SketchUp application, and plans drawn up to repair the damage. The metaview afforded by the 3D model allows the viewer to experience something that has been lost, to get a sense of the place as it was before a destructive event. The purpose is not only to help give viewers an appreciation of the value of the structure in its environment, but to provide the information necessary to support a reconstruction project.

In this case, despite several year of effort before the fire to collect them online, sufficient photographs of the hut were not immediately available to create the 3D model, but a call-to-action among KHA members immediately after the fire saw sufficient photographs for the 3D model to be made contributed to the KHA photo website. Had the photographs scattered across the internet in, for example, KHA members’ Flickr accounts, on Panoramio or in the online collections of cultural institutions, contained location metadata the call-to-action may not have been required.

[INSERT PHOTO3 HERE]

Whites River Hut, Kosciuszko National Park, February 2010 [ Photo: James Steele]

Finding geotagged photographs easily remains difficult. A Google Image Search for Whites River Hut returns 1.35 million results but, in fact, few of the images are relevant. If, for instance, these images were geotagged, it could be possible to search just for images within the immediate vicinity of the hut, and building the metaview would have been a much easier project.

It is clear from examples such as these that the technology to search images by location, essential to the creation of a metaview of a particular place, is still nascent. For instance a name-based search for a location on Trove returns not only photographs, but a range of other sources including books, journals, magazine and newspaper articles, oral history recordings and maps, not all of which may relate to the particular location under investigation. What does not yet exist is an effective search engine for images from across the internet that returns results based on the unique location where the image was captured. Online photo collections services Flickr, Panoramio and Picasa have map-based interfaces to their own collections that allow visitors to browse for photographs taken in particular locations, but as yet there is no over-arching search engine that would return all the photographs from a particular location across all three services and the rest of the internet at once. Limiting the results of a Google Image Search by location would be an important feature, but it requires the images to have the necessary and accurate metadata encoded along with the image; an ability for the search engine that located and interpreted the image file to understand the embedded location data; and an interface to allow the user to limit their search to a particular location, or to have the results grouped by location if the searcher was unsure of where the place they were searching for actually was.

Building new means of navigating the found set of co-located images is also a challenge. Microsoft’s Photosynth provides a better experience of a place than the scattered icons and picons of Google Earth’s Panoramio layer. However the experience is still one of moving between the flat, framed rectangles of the original photographs that have been digitally stitched together into the synthetic view. In one sense this effect is a salutary, but constant, reminder for the user of the origins of the view as disparate photographs. However a more seamless, immersive, and perhaps visually exhilarating experience of a place could also be generated through the amalgamation of games technologies like Epic Games’ Unreal Engine which could recreate real-world environments in 3D with photographs harvested from the internet. [25] [26]

In any case crowdsourced metaviews, as they become more accessible in the future, potentially offer a much richer experience of the place, drawing on different points of view of different visitors at different times. These representations may provide resonances that enrich the viewer’s experience of a place beyond the framed, well-executed, single image. Even now experiencing Uluru via Google Earth allows a far more complex and complicated interaction for the casual viewer than that available through the iconic but flat picture postcards or Australiana picture books. There is the potential through this sense of ‘virtual fly-through’ to provide the user with an exhilarating sense of access to significant places that is a more materially based experience than the iconic image can ever provide. It may even replace the commanding monocular imperial gaze with one that is more engaged and enmeshed in the place.

[INSERT PHOTO 4 HERE]

Uluru on Panoramio[27]

However, currently iconic views of iconic places still dominate Google Earth’s metaviews. Maps of the distribution and concentration of photographs on Google Earth demonstrate that the Opera House, Bondi Beach and so on are all places where the crowd still gathers to take their own photographs, just as the professionals have stood there before them to make their postcards. However metaviews which incorporate additional images taken from different perspectives, at different times of the day, showing fleeting visitors or transient events offer at least the possibility of an enriched understanding of these already famous places. Other places, for instance those shrouded by difficult terrain, property laws, prohibitive cultural taboos and so on may also become accessible through metaviews when enough sufficiently tagged images become available.

This may take some time however. There are still many factors restricting the scope and scale of geotagged images being collected and navigated. For instance Panoramio, the source for Google Earth’s picture layer, has strictly enforced rules about what can and cannot be contributed to the site, and what will or will not be selected for viewing on Google Earth. Actual people manage the selection process manually, and the rules exclude (with some exceptions, interpreted by the people making the decisions), among other things: portraits of people, or photos of people posing; cars, planes or any machine; pets or animals; flowers and details of plants; close-ups, details, inscriptions, or signs; events, such as exhibitions, concerts, and parades; interiors; anything under a roof. [28] These exclusions prevent many examples of culture from being displayed alongside other images of a place. The excluded images may contain evidence of activities and details that help viewers interpret the place. While Google (which owns Panoramio) has every right to set its rules for users of the site and to control the process by which images are displayed on it, the rules and processes limit the opportunities for the service to enrich viewers’ experiences of the place. By its restrictions on what can and cannot be displayed, Google Earth is not a service where place can be explored. Evidence of space and location are available, but not the human activity and artefacts that make a space a place.

The metaview has the potential to give us a new sense of place if we can overcome the barriers to images being contributed and made available to the community. Crowds are capturing images with digital cameras and libraries are scanning older images, but we are not yet sharing them easily. Perhaps we are afraid of being ripped-off or the images being somehow misused. Metadata is being added to images, but it needs to be more accurate, more extensive, and using generally agreed language and standards. Search engines continue to improve and provide extra dimensions to finding photographs that exist on the internet, but the display of the results largely continues to be limited to outdated and two-dimensional, using page, document and book metaphors and logics in their interfaces, rather than image based or scopic metaphors and logics.

And of course the geopolitics of wealth, religion, culture and class define the geography of World Wide Web, just as they define our real world. As is so starkly revealed by the world heat map of photographs uploaded to Flickr, the technologies that are being developed to record the images, store them, share them and display them, and the people with the resources to access the technologies and the skills required to use them, are the richer and more educated people of the world. It is they that determine the direction of the technologies and the uses to which they are put, defining the perspective we take through the metaview.

Nonetheless it is clear that we are at the threshold of the age of the metaview. In Australia, libraries and ordinary people are beginning to collaborate in an unprecedented way to allow us an enhanced scopic access to Australian places. What remains to be seen is whether or not these new modes of experience will change the nature of the places we collectively decide to value as our national heritage. Will it continue to be those places that are most amenable to being encapsulated into a single frame as a vista, or will other places, where perhaps multiple layers of experience lie embedded, also now rise to national significance?