Module 2 - Visual Encodings and Color Flashcards
Marks (V1)
Geometric primitives
a) Marks for Items
- Points (0D)
- Lines (1D)
- Area (2D)
- Volume (3D)
b) Marks for links
- Containment
- Connection
Visual Channels (V1)
Control appearance
- Position
- Size
- Color
- Shape
- Angle
- Motion
Visual Encoding (V1)
Combination of marks and visual channels to show data attributes
Principles for visual encoding? (V2)
- Expressiveness Principle
- show all that is connected to data
- match channel/mark to data characteristics - Effectiveness principle
- Encode the most important attributes with highest ranked channels
Channel Ranking (V2)
a) Categorical Data
1. Spatial Region
2. Colour Hue
3. Motion
4. Shape
b) Ordered Data
1. Position on scale
2. Length
3. Angle
4. Depth/Area
5. Color
6. Curvature/Volume
What are the channel rankings based on? (V2)
- Accuracy (percieving changes to same channel)
- Discriminability (how many values can distinguish and compare)
- Separability (interaction between different channels)
- Pop out (parrellel processing)
Facts about Accuracy (V3)
- Humans better at measuring length then angle
- Saturation overestimate, Area under estimate, Length is best
Stevens Pyschophysical Power Law (V3)
for visual channels, relationship between percieved sensation and physical intensity
S = I^N
S = percieved sensation
I = Intensity
Perceptual Scaling (V3)
In order to counteract bias, use power law to scale the data (for the sake of perception for users)
Discriminability (V4)
- How many values are we able to distinguish and compare?
- Precision match encoding
Separability (V4)
- Be able to identify attributes, without interference
position hue are fully seperable
size hue some interference
width height significant interference
mixing hues full interference
Pop out (V4)
Pre Attentive
- Color or Shape Individually
- Parallel Processing (items seen immediately)
Not Pre-Attentive
- Color and Shape Together
- Serial Search (speed of search depends on distraction count)
How do we see (V5)
Reflection and Absorption of light on objects
- 400 to 900 nanometers wavelength of electromagnetic waves (blue= 400, red = 900)
- can see about 10 million colors
Rods (V5)
Help with “night vision” (Achromatic Perception)
- do not see color
- saturated through the day
Cones (V5)
gives color to images
- 3 types of cones: short (blue), medium (green), long (red)
Color Blindness (V5)
Deficiencies of functioning cones
-> Red green color blindness (5-8%)
-> Blue yellow color blindness (1%)
How do we represent color digitally (V6)
Through physical realisation
- screens use RGB Space, 3 dimensions (since 3 cones)
- printers use CMY
Semantic of Color Representation (V6)
hue = color wheel
saturation = how much grey (power of the colors)
light = how much brightness
CIE 1931 XYZ (V6)
mathematical model used to represent ALL human colors. absolute color, NOT device dependent
- device dependent models are mostly parts of the CIE 1931 XYZ space
Perception facts (V6)
Better at detecting/distinguishing green
Blue perception is weak
Very sensitive to luminance (light intensity per unit area)
Visual semantic to represent categorical data (V7)
Hue, max 12 attributes to distinguish colors well. use constant luminance
Visual semantic to represent ordered, sequential data (V7)
luminance/black body color maps
-> hue is NOT a good choice since it is not naturally ordered
-> higher accuracy in percieving luminance then HUE
(can distinguish 100 shades of grey, but only 12 colors)
Visual semantic to represent ordered, diverging data (V7)
colormaps
-> dont use rainbow hue colormaps, instead opt for two distinct colors
Webers Law (V8)
Perceptual system mostly operates with relative judgments, not absolute. colors,size,luminance are not evaluated independently, our brain uses information from experience AND context.
change in s
—————- = k
s
the relation between stimuli and detecting change is always constant
(more stimuli in environment = more change needed to notice there is change)
