#csaunders

Google Filter commenced in filming a large-scale LED advertising screen in an effort to understand the nature of the color changes emitted from moving light images. I sampled film stills from the footage at points where color differentiation was at its most acute, but my systematic approach of addressing the same point within each still did not produce compelling results. The differences in color throughout various structures within the composition existed in dynamic interplay with one another, and one point was not a measure of the entire effect. Therefore I began submitting the images to Google’s ‘search by image’ function, and it was at this point that I was able to understand the change in color tone as the screen progressed over time, as well as the ratio of one color to another within a single frame. 

This, of course, is a description of color from the point of view of a computer algorithm. I am intrigued by the images output by Google – often they capture nuances that belie the work of the computer at hand, occasionally it produces a whole set of equivalents that faithfully capture the essence of my input without any similarity of content. 

‘Cambridge’, on the website, is an example of a Google formulation: an algorithmically contrived interpretation of a filmed city walk, whereby a frame selected every ten seconds is input to image-search, resulting in myriad variations of a given color scheme, form pattern, or shape association reference.

The most recent addition to the project is the inclusion of Google’s word search mechanism, thus adding a verbal component to the imagery if the original material allows. As I transcribe into the search engine, I record word structures that are compelling in their implicit ubiquity – they would not have been sourced had they not been previously searched. By lending Google the autonomy of ‘choosing’ images and sentence fragments, the engine begins to formulate a distinct narrative.

 To assert the symbiotic value of image and verse it has been important to select from conventionally produced filmic works that incorporate both mediums. The decisions that were made in creating the qualitative product of a film are important in garnering robust output imagery, and the contrived narratives I sample are deliberately framed, colored, and imagined. I have employed perfume advertisements – Chanel No. 5 (1983), Thierry Mugler Angel (1993), L’Eau D’Issey (2011) – because they are short, possess several different shots or scenes, are artfully administered, have dialogue, and are reflective of a niche society.

To communicate that which is intangible has been the objective of perfume commercials through the decades. Brands invest a significant portion of their advertising allowances to deploy exaggerated imagery and melodramatic narrative that is often articulated through voices and phrases which might ring despotic in their fervent promotion of product.

Because it is impossible to convey scent through audio-visual media, the content of the films become an essential visualization of the brands’ intended perceived message: the smell has become the brand. These short films – often between thirty seconds and one minute – are highly controlled productions. The frame of a still, the distribution of light, and the selection of colors present within an image are specially designed to reinforce the product and brand at hand. The results, then, are direct reflections of the meticulously researched desires of the contemporary culture at hand.

I present a selection of commercials representing well-known perfumes with the intention of exploring their visual and semantic data through the by-products of Google search filters. By distilling frames and linguistics into separate blocks of information, the aggregate elements that define a cultural index can be precisely represented. The subsequent recombination of the meta-data sets allows for discordant and diverse reinterpretation.

Ultimately I found the visual documentation of the city to produce the most fruitful rendition of Google Filter. I plan to continue documenting cities in this format, and revisit selected images over time to measure differences in Google’s output as their image database constantly changes – perhaps the project can unravel as a small study of the evolution of the search engine itself.

www.google-filter.biz

We've had enough with unintuitive calendar apps that mimic the traditional paper version. It's time for them to think at our speed and embrace creative interfaces. Introducing CALENDAR.

http://prezi.com/x6ifmdt8oowu/es22-c4/  

(video is embedded into the PREZI)

Returning to South Station, I filmed various sections of ceiling that reflected the light emanating from the LED advertising screens. It was slightly earlier in the afternoon than my initial data collection session, and I was worried this may affect the outcome. Comparing the two segments, little changed. After uploading the recorded material I paused each time the light changed drastically. Often a shade of light would last as little as a second or two, and would change in intensity momentously. I captured film stills as these segments changed, and aligned them chronologically. In order to ascertain the change in color, I chose a specific place in the shot to be the control – a ceiling beam which jutted from behind the screen at about mid-level, an element of each image that appeared be affected quite drastically by the light.

Using illustrator’s color sensor, I sampled the control point in the consecutive film stills. Disappointingly, the outcomes were less varied than I expected: mauve, brown, grey, black. This demonstrates the dynamism of an image as being a result of the combined differences of multiple planes – the perception of single point does little to impact one’s cognition of an expanse of space. The color sensing mechanism of an iphone is, of course, different from that of the human eye, but for now it is the best measuring tool I have.

I wanted to convey the data as Ed Ruscha did in his ‘Every Building on the Sunset Strip’, 1966. Perhaps I might use his format as influence to expand upon my collection of data, since I am underwhelmed by the results. Instead of using the meager screens of South Station, I am considering the environment of Times Square, it being a selection of varied screens and colours. The exterior environment may complicate establishing a point of light reflection, and certainly a control point will cease to be since the focus will be on the various facets of this much larger space. It is hoped that a camera can capture the glow of a light, a colour of light, in immediate proximity to its screen.  

Indeed if in a different location the data would evolve into a divergent set of observations, albeit with the initial intent still intact. If it were to remain in Boston, are there other locations that LED screens inhabit?

I also recorded the circuit of advertisements in South Station that I was watching, and the time of each, as follows:

Tropicana: 10s

Equitylease: 10s

South Station: 10s

Bruins: 30s (only one with narration)

Tropicana: 10s

Blue Man Group: 30s

Museum of Fine Art: 15s

Advertising Opportunity: 10s

me: no?

my brother: Then I'm not interested

me:

me: you're 10

me: u

They could hypothetically break all of our hearts and make Cam commit suicide. I'd cry for days, but they definitely have the opportunity for that.

Oh God.