Visualization Design on Large, High-Resolution Displays
by Andrew and Beth
Perception: Are graphical encoding orderings different on large/high-resolution displays?
1. Spatial Position effected by
a) physical size of display:
- causes outer spatial locations to be in peripheral vision
- need to maintain physical context
b) DPI combined with physical size
- presses limits of visual acuity
c) bezels
- can cause spatial distortion
- can also hide data behind the bezels
2. Length effected by
a) physical size of display
- hard to compare lengths that are far apart
b) DPI
- if length is based on number of pixels, higher DPIs lead to smaller differences
3. Color effect by
a) display brightness, contrast, color calibration, etc.
- colors look different across tiled displays
b) viewing angles
- based on the quality of the display, the viewing angle can affect the color encoding
4. Orientation effected by
a) physical size of display
- viewing angle can change apparent orientation of a line
Visual Metaphors and Scalability: Which designs and datasets benefit most from these displays?
1. Which types of data most benefit?
a) geographical applications benefit because more geography is shown at once
b) bioinformatics, file systems, social networks, most visualizations
2. Benefits for different visual metaphors depend on fix-ability of scalability problem:
a) fixable issues
- number of bars
- glyph size
- aspect ratio issues
- over plotting
- distortion
b) non-fixable issues
- number of perceivable colors
- 3D occlusion
- network connectivity over plotting
3. Cognitive affects of scaling visualizations on these displays:
a) Certain visual metaphors are likely to be more understandable when scaled
b) Some visualization techniques can become confusing (more edges in a graph = more intersections)
c) May find more local patterns on a large display and global patterns on a small display
Interaction: How does interaction with a visualization change on large/high-resolution displays?
1. Trade virtual navigation for physical navigation
2. Overview techniques change:
a) may be able to display all data rather than eliminating data or aggregating data
b) visual aggregation may become more prevalent (or used to aid computational aggregation)
c) still needed overview strategies because
- dataset may still require more pixels than available on large displays
- cognitive or perceptual problems are possible when huge amounts of data are shown
3. Navigation strategies change:
a) overview + detail
- different zoom factors may be necessary
- cascading overviews are possible and perhaps necessary
- must decide where to put the overview
- must decide how much display space to use for overview(s)
b) pan+zoom
- scrollbars are impractical for panning on large displays
- must be able to change the zoom center point to make use of the entire display
c) focus+context
- may be useful for collaboration on large displays: fisheyes combine
d) physical versus virtual navigation
- physical navigation has advantages in speed and maintaining context
- virtual navigation requires less effort from users
4. Different brushing and/linking techniques may be necessary:
a) problems:
- change blindness may become a major problem
- may not see changes in the periphery
b) solutions:
- potential to use motion, but may become too distracting
- could base update on when the user looks at a view
- may want to change the speed of updating across views
5. Trade-offs in having controls on/off a large display
a) controls independent of display:
- mobile; user can walk around, remain stationary, and view any area of the display
- ui/controls always with the users
- controls are consolidated into a central location
- users must carry around the external device
- device can’t take advantage of the benefits of such a large display
b) controls/ui on the display
- user does not have to use excessive external devices
- the controls can benefit from the screen real estate
- controls can become distant from the user, hard to access
- the UI generally does not scale from 1 screen to display walls