Visualizing 3-D data from a static 2-D perspective projection imposes limits on the viewer's ability to intuitively understand the data. Without further interaction or rotation, the user must surmize depth information from perspective cues. Furthermore, if the data has many dimensions beyond physical location, there is a need for other visual methods to be employed to capture these other aspects. Color is an easy solution for one field, small glyphs in a regular ordering can aid a few other's, and contour lines often are unobtrusive. However, the combination of all these visual elements can result in a densly cluttered image that will require significant attention and processing in order to fully understand all the data it portrays. The key insight here is that the visual modality is being taxed heavily because of its broad capacity to interpret many different types of stimulation. But, as more components are placed next to eachother, it becomes difficult to percieve them as relating to a unifying field.
As long as user interaction is allowed, a solution to this is to add another perceptual modality to the data representation. The most feasible second sense available to a human user is hearing. Like graphics, computers are able to produce high-quality audio output at roughly the same (or a lower) cost. Also, speakers are a ubiquitous part of modern computer hardware.
This project will use a combination of visualization and sonification to enhance the user's ability to percieve the data from a complicated flow field. The data to be used will be the weather data from Hurricane Isabel (program 1). This multivariate data set has dimensions like wind speed, wind direction, temperature, pressure, and an array of information about different kinds of water states and masses. Some of these dimensions have simple correlates in perception, for example, red things in nature often occur in a red color where as cold things appear blue, thus, mapping temperature to color seems intuitive. Also, powerful movement is often accompanied by loud noises, so magnitude of a flow vector can be mapped to volume. Pressure can be thought of like a kettle that fills with steam until the pressure is too great and the steam bursts from the nozzle as the kettle screams a high note, thus pressure can be mapped directly to pitch or frequency in the sonification. Direction is a topic that is only suitably represented as a visual component, since it involves spatialization and implied motion, which are things observed by sight. There is still room to experiment with novel connections between data and perception, but just these initial linkages are enough to justify this endavour.
Picking with the mouse to hear different sounds involved in the sonification, ability to rotate the 3D data field so that new areas of sound can be explored.
| Week |
Work Goal |
Week |
|---|---|---|
| Five |
Understanding/parsing the raw data, viewing the data with paraview for reference, reading the data into a program, and organizing the data in a way that is functional to work with at the code level. |
Five |
| Six |
Visualize the data using psuedo-coloring, contour lines, and arrow glyphs. |
Six |
| Seven |
Get sound into the code for sonifying the data using simple synthesis with parameters from the data. |
Seven |
| Eight |
Link visualization with sonificaiton and fine tune the user interaction component. |
Eight |
| Nine |
Debug, add any necessary documentation to the code, collect media and sources for the final webpage, and write a rough draft of the written report. |
Nine |
| Ten |
Finish the final draft of the written report, format the webpage, and practice presenting. |
Ten |