| torifoster.com | tori@torifoster.com

 
   
  The Mother City
 
In progress documentation
 

This page will be updated on an ongoing basis with the progress of my MFA Documentary Media thesis project.
My thesis advisors are Alex Bal and Steve Daniels. David Green has also been of great help with this project.
Please contact tori@torifoster.com for further details.
   
Project Description
The word metropolis has its origins in the Greek metros (mother) and polis (city)—a city whose characteristics are representative of its nation. If this is true, what do the characteristics of Canadian and American cities say about their respective nations. What do the competing urban ideas, made concrete by architecture and planning projects, reveal about the ideologies and ways of life of our societies? By documenting and representing the physical forms of these cities my aim is to express the individual characteristics of cities, as well as to determine whether there is an archetypal structure common to each of the nation's metropolises.

The Mother City is a new media typology that represents the section of 22 different cities in the United States and Canada. An installation incorporating sound and video will both document and reconstruct the spatial environment of the metropolis. Viewers will witness discrete typologies displayed around the circumference of the gallery which depict individual cities, as well as a composited typology at the gallery centre that constructs the archetype of our cities' forms.
 
Installation scheme
Click image to see in full.
________________________________________________________________________________________________________________
August 21, 2008 List of online references for:
The Mother City documentary
The Mother City thesis paper
   
________________________________________________________________________________________________________________
October 8, 2008 Data collection routes:
X  Toronto
    Ottawa
X  Montreal
    Quebec
    Buffalo
    Rochester
X  Boston
    Providence
    Hartford
X  New York
X  Philadelphia
    Baltimore
    Washington
    Detroit
X  Chicago
    Milwaukee
    Indianapolis
    Louisville
    Cincinnati
    Columbus
    Pittsburgh
    Cleveland

Those marked with an X are the 6 cities that I will be beginning with, and adding to if necessary.
   
________________________________________________________________________________________________________________
July 15, 2008 Figures A, B, C and D represent display strategies for The Mother City.
  Figure A: The ghost composte: discrete city sections overlayed upon each other transparently.
   
                                       
  Figure B (left): Two strategies for typological study:vertically (full city section), and laterally (city zones).
   
  Figure C (right): Another representation could take the form of a ‘sampled' city, where each typological category (power centre, suburbs, high density residential, downtown core, etc.) is the constant but every 100 metres is represented by a different city. Thus you might see the Empire state building (New York) next to the CN tower (Toronto) which is next to the Sears Tower (Chicago). You might also find six Home Hardware stores side-by-side.
 
________________________________________________________________________________________________________________
July 21, 2008 All images below were generated after doing a test run down King St., Toronto, on July 21, 2008.
The purpose of these tests is to find the best sectional representation of the data I've collected.
Images are displayed in chronological order of production.
  Figure 1: North-west corner of King and Bay looking north. This is the first image created using Processing. It is a composite of translucent stills taken from the bottom video camera (which captures parallel to the horizon).
   
 
 

Figure 2: A quick look at the top camera's output (perpendicular to the horizon). Looked promising to begin with but yielded less than promising results when more sample images were used (see figure 3 below).

   
 
  Figure 3: Distortion (described in figure 6).
   
 
  Figure 4: After trying Photoshop CS3's new stitching software (which failed to produce any result), I tried PTGui software. As you can see here, it worked splendidly with a sample of approximately 8 images.
   
 
  Figure 5: Unfortunately the same software (PTGui) yielded less than desirable results with a sample number of about 40 images. This software is predicated on a stationary viewpoint, which I never have while shooting.
   
 
 

Figure 6: When compositing by drawing one image on top (and slightly to the left) of the last, the right side of each sample image is privileged because that side of the image is closest to the top of the layers (and therefore can be seen most clearly). As such the pull towards the centre (caused by the barrel distortion of the wide-angle lens) is what is most apparent in this composite.

   
 
  Figure 7: In an attempt to remedy the left tilt (seen in figure 5) I tried using only the central region of each sample image. The result is more analogous to the environment it represents.
   
 
  Figure 8: Detail of figure 7 above.
   
 
 
  Figure 9: This image uses each sample image in it's entirety after a filter (left and right edges fading to white) has been applied to each sample.
   
 
  Figure 10: This image was generated using a mask that makes left and right edges of sample stills fade to transparent. This eliminates the vertical striations seen in figure 9.
   
________________________________________________________________________________________________________________
July 30, 2008 The two images below were generated after doing a test run down King St. , Toronto , on July 30, 2008.
The purpose of this test was to see if using one fisheye lens (Raynox QC303) would better capture my subject. This was in part executed under the pretense that the fisheye lens would yield a 160 degree angle of view, where in fact it was more like 45 at best, from top to bottom.
   
 
  Figure 11: In this image you can see as high as the top of First Canadian Place and as low as the second floor. The this image is generated from one the captured contents of one lens (with a fisheye attached). First Canadian Place is drastically foreshortened--so much so that the upper-most 72nd floor looks like the third floor.
   
 
  Figure 12: Detail of figure 11.
   
________________________________________________________________________________________________________________
July 31, 2008 The images below were generated using the test run completed on July 21, 2008.
   
 
  Figure 13: This is a panorama of King St. focusing on First Canadian Place. The image contains the panoramas generated from all three cameras ('top' centered at 90°, 'middle' centered at 55°, and 'bottom' centered at 20°).
   
 
  Figure 14: This panorama merges the three streams in Photoshop subsequent to them each being drawn in Processing.
   
 
  Figure 15: Detail of figure 14.
   
________________________________________________________________________________________________________________
August 18, 2008 Prototype camera rig.
   
 
  Figure 16: Prototype rig from exterior of car. (Two wide-angle lenses are not seen in this image). Needs cross braced reinforcement on sides, as well as ridges on camera seats to prevent rotation.
   
 
  Figure 17: Rig from interior of car (view towards passenger window). I also didn't anticipate the tilt upward that would occur when the metal was lashed to the door handles, which caused all the camera angles to decrease by 10° (45° angle became 35° etc.).
   
________________________________________________________________________________________________________________
June 15, 2008

<trkpt lat="43.629943" lon="-79.424855"> <ele>96.269287</ele> <time>2008-07-15T14:08:12Z</time></trkpt>
<trkpt lat="43.629989" lon="-79.424605"> <ele>91.462524</ele> <time>2008-07-15T14:08:16Z</time> </trkpt>
<trkpt lat="43.630031" lon="-79.423996"> <ele>87.136597</ele> <time>2008-07-15T14:08:20Z</time> </trkpt>
<trkpt lat="43.630087" lon="-79.423044"> <ele>87.136597</ele> <time>2008-07-15T14:08:25Z</time> </trkpt>
<trkpt lat="43.630219" lon="-79.421787"> <ele>84.733398</ele> <time>2008-07-15T14:08:31Z</time> </trkpt>
<trkpt lat="43.630417" lon="-79.419687"> <ele>81.849365</ele> <time>2008-07-15T14:08:41Z</time> </trkpt>
<trkpt lat="43.630679" lon="-79.417950"> <ele>78.965454</ele> <time>2008-07-15T14:08:49Z</time> </trkpt>
<trkpt lat="43.631060" lon="-79.415596"> <ele>77.042847</ele> <time>2008-07-15T14:08:59Z</time></trkpt>
<trkpt lat="43.631395" lon="-79.414023"> <ele>76.562134</ele> <time>2008-07-15T14:09:06Z</time></trkpt>
<trkpt lat="43.631834" lon="-79.412514"> <ele>78.004028</ele> <time>2008-07-15T14:09:13Z</time></trkpt>
<trkpt lat="43.632234" lon="-79.411235"> <ele>82.810791</ele> <time>2008-07-15T14:09:19Z</time> </trkpt>
<trkpt lat="43.632283" lon="-79.411038"> <ele>82.810791</ele> <time>2008-07-15T14:09:21Z</time> </trkpt>
<trkpt lat="43.632285" lon="-79.410995"> <ele>84.733398</ele> <time>2008-07-15T14:09:24Z</time> </trkpt>
<trkpt lat="43.632310" lon="-79.410944"> <ele>83.771973</ele> <time>2008-07-15T14:09:51Z</time> </trkpt>
 

GPX track data retrieved from a trip taken June 15, 2008. You can see that the time stamps average every 4.5 seconds (highlighted in green). Also, stamp intervals are not consistent. At times the data is recorded as wide as 10 seconds apart (highlighted in dark green). Traveling at an average speed of 50km/h, an average of 4.5 seconds translates to data being recorded every 62.5 metres.

In order to use the GPS data as an editing tool, I need to have data recorded, at the very least, every 3 metres. I later found a setting that allowed me to record data not based on time, but on distance traveled, the lowest tolerance being every 10 metres. Below is that test. However, most likely the failure will be that the GPS unit is only accurate to within 3-7 metres.
   
 
  Figure 18: GPX track from June 15 visualized using 'GPS Visualizer' online mashup application. Right now I see the GPS uint serving this auxiliary function--a key diagram of the section (such as figure 14) displayed beside/proximal to each photograph.
   
 
  Figure 19: GPS track from August 21 visualized using 'MapSource' using a track that stamps every 10 metres. Generally the GPS successfully plots every 10 metres, but occationally plots every 20 metres. (From red point to orange is 10m, orange to blue is 20m).
   
________________________________________________________________________________________________________________
Ocotober 7, 2008
  Figure 21: Example of width and height measurement.
   
 
 

Figure 22: While pulling routes on Google maps, I came across these two suburban conditions. The houses that line the arterial road (seen here second from the right) turn their backs to the road, rather than the previously traditional front facing orientation. Also sidewalks are reduced from two to one.

This has an impact on choosing routes. In order to obtain an accurate reading of the elevation intended for the public I'm going to have to intentionally deviate from the otherwise more direct routes while travelling through suburbs of this nature.
   
 
  Figure 23: I'm currently retrieving the city limits for each of the 22 cities to be documented (New York pictured here). The strategy for the driving route is currently to drive from city limits 'A', to the tallest building, to city limits 'B' (choosing the two points at city limits with the greatest distance between them). The route will incorperate _____ (number) of attributes described in each of Le Corbusier's and Wrights' conceptual elements of a well functioning city.
   
________________________________________________________________________________________________________________
October 8, 2008
  Figure 24: Cities to be documented in The Mother City as of October 8, 2008.
   
 

Safety: As I've been tracing driving routes through each city, I've had to confront the issue of safety while driving through dangerous neighbourhoods in certain American cities. One solution to the problem of safety is to just omit the unsafe cities and go to only cities that will not cause safety issues. Another is to look atradiating lines of Toronto, doing an analsis of only Toronto. I'm currenly leaning more towards the first solution, so I'm now trying to find out which are the cities to avoid.

   
________________________________________________________________________________________________________________
October 22, 2008
  Figure 25: I've been saying for quite a while now that people are not represented in this project--but after taking a second look at my test images it is clear that they are in fact represented quite clearly. Click the image to see in greater detail.
   
 
  Figure 26: Installation mockup. Changed as of October 29, 2008. After re-evaluating my thesis statement I found that this installation scheme did not deliver an appropriate environment in which to concieve of the metropolis in its entirety. I imagine that it would in fact reinforce the problem of fragmentation given the inablity to see each of the monitors in the same field of vision.
   
  Figure 27 : Installation scheme, corresponds to figure 26. The installation environment will have 22 LCD monitors arranged side by side, spaced evenly around the circumfrance of the space. One speaker will be placed above each of these monitors. Each speaker, and the monitor below it will depict a representation of a city.

 

  
________________________________________________________________________________________________________________
November 6, 2008
 

Figure 28 & 29 : I'm currently trying to output my sketches to .mov files but I'm having difficulty with the resolution in the videos. The problem is a strange one though: the sketch looks great, the .mov file on my computer doesn't, but when I uploaded it to my website and played it from that server it looks ok. See that video here.

 
   
________________________________________________________________________________________________________________
November 8, 2008
 

Figure 30 : This is the beginnings of a device that will measure location with time stamps. The white plastic cases at top house a magnet and a reed switch. The magnet will be attached to the wheel of the car, and the reed switch to the undercarriage. Each time the magnet passes the switch the switch changes states (from open to closed). This signal is sent through a processing sketch that will in turn record (to a text file) the time of day at which the switch changed states. The information gained will look something like this:

0 metres: 10:00:00 AM
1 metres: 10:00:01 AM
2 metres: 10:00:04 AM
3 metres: 10:00:09 AM
4 metres: 10:00:20 AM
5 metres: 10:00:34 AM
6 metres: 10:00:60 AM
7 metres: 10:00:98 AM
8 metres: 10:01:51 AM
9 metres: 10:02:20 AM

This information will be placed into a second processing sketch. Instead of drawing the stills at a regular time interval, the sketch will draw the stills at regular metre intervals. Thus, the resultant image is one that privileges space over time (the buildings rendered will be of an accurate spatial scale).

   
________________________________________________________________________________________________________________
November 9, 2008
 

Figure 31: I am contemplating adding human experience to the project by tracking the spatial movements of Torontonians. One of the initial motivating factors in creating this piece was to give opportunity to the individual to conceive of the metropolis as a whole. The metropolis is a monster in its size and I believe that most metropolitans experience only a fraction of the city they inhabit. A repercussion of this is that one's understanding and mental picture of our urban cities is murky at best.

Further, why is it that these limited spatial narratives exist? Is the metropolis too great a scale for human living?

   
________________________________________________________________________________________________________________
November 10, 2008 Merged Video Streams I've successfully created a processing sketch that can export to video three stills streams, stitched both horizontally and vertically. Click here to see the video.
   
________________________________________________________________________________________________________________
November 20, 2008 Digital Painting I'm now conceiving of the final product not as an installation but rather as a 'digital painting.' This is an object that can be picked up and hung on a wall... fall off of a wall, taken off the wall, and moved to another, all as a single object. The object would be plugged in. That's the difference. I'm working on a mock up of this. More to come shortly.
   
  Aliases Naming conventions for the citizen aliases will be derived from the influential architects urban planners and critics of the 20th century (Frank, Jane, Ebenezer, Pierre, etc).
   
 
 
 
 
  Please contact tori@torifoster.com for further details.