32 2. FIVE STORIES TO A MODEL OF VIDEO STRUCTURE
marathon videos, the camera was often on a motorcycle alongside the runners; potholes in the road
would cause minor variations in the camera/object relationship.
2.3.5 RESULTS
e salient gures from the analyses of the two marathon videos showed 3:1 ratios of longest image
set lengths, range, and mean lengths, indicating very dierent structures. While Document A is
only 58% as long as Document B, it is made up of 174% as many image sets. In a similar manner,
the runners in Document A typically occupy nearly twice (196%) the screen area of those in Doc-
ument B. In the current data Object 2 represents the average area of an individual runner within a
recorded group of runners. e near 2:1 ratio holds throughout the documents.
Some 90% of Document A image sets lie within the 10-second range, with a rapid decline
to the single longest image at 19 seconds (17–20). Document B, on the other hand, presents a less
dramatic rise and fall and has nearly 30% of its image sets in a length range where Document A
has none at all. is means that in Document A we are presented with faster paced editing (more
image sets of shorter length) and that the women pictured are larger on the screen.
ere is no value judgment implicit in the data yielded by the analysis of the two videos.
Even if the hypothesis that rapid visual processing is more enjoyable, all else being equal, we must
remember that all else is often not equal. e contour map indicating rapid change in data does not
say: is is better. e contour map indicating slower pace does not state: is is dull. A topographic
map of hill terrain does not make an evaluation of hill compared to plains. A weather map with
closely spaced isobars does not say: Windy weather is better. A running coach who viewed the two
documents pointed out that if he wanted parents to be excited enough to donate money to buy new
uniforms, he would show Document A; yet if he wanted to show his runners how to strategize and
how to “read” the other runners in a race, he would show Document B.
2.3.6 KEY FRAMES AND STRUCTURE PATTERNS
Tracking continuities of attributes and seeking points of discontinuity provides a means of rep-
resenting document structure, and, thus, the foundational material for abstracts. Frames on either
side of a point of discontinuity are key frames. Key frames and structural attributes are the princi-
pal entities used to reveal something essential about a moving image document. ey are smaller
quantities which make available the virtue and power of the document in a medium-specic man-
ner. ey are the boundaries of dierences that make a dierence. e data entered into the test
environment program for any particular document can be used for numerous sorts of analysis and
access. Individual frames, sets of frames, and alphanumeric and graphical data derived directly from
the frames are used to construct a variety of abstracts.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.