6. Marks and Channels

Why marks and channels?

Every complex visual idiom can be broken into two orthogonal(?) components

→ Marks and Channels

Marks

• Basic graphical element.
• Classified with their spatial dimension
  • Point, Line, Area, Volume
  • Actually, area and volume are also channel itself. (그럼 orthogonal 아닌거 아닌가…하지만 넘어갑시다)

Marks as Items/Nodes

• Points
• Lines
• Areas

Marks as Links

• Containment
• Connection

by Jacques Bertin

---

Channels

• A way to control the appearance of marks.

• human perceptual system has two fundamentally different kinds of sensory modalities

  • Identity (저게 뭐지)
  • Magnitude (양 비교)

• Position

  • Horizontal, Vertical, Both

• Color

• Shape

• Size

  • Length, area, Volume

• Tilt

So… Marks and Channels again?

• Marks → geometric primitives
• Channels → control appearance of marks
• Interactions (they are dependent each other) (아니 그럼 더더욱 orthogonal한게 아니자너) → Shape and Size channels cannot be used
  • Area marks are fully constrained. (state or province…)
    • Cannot be sized or shape coding.
  • line mark convey position and length.
    • can only be size coded in 1D
  • point marks only convey position.
    • can be sized or shape coded.

Note: this is vertical:position channel with line mark.

---

Note: People gaze at marks’ height

---

Expressiveness and Effectiveness

(3번째 나오는 중)

• Expressiveness

  • Vis idiom should express all and only information in the data attributes.
  • match channel and data characteristics (identity channel, magnitude…)

• Effectiveness

  • Most important attributes should be encoded in most effective channels
  • ex. Vis → Task → 5 Attribute가 제일 중요하다 → 중요도에 따라 effective channel을 Mapping 해야한다.
  • 그럼 그 중요도는 어떻게 순위를 매길 것인가
    • Accuracy
    • Discriminability
    • Separability
    • Popout

  ---

  즉, expressiveness는 왼쪽 오른쪽 타입을 data type에 잘 맞추라는 거고,

  effectiveness는 중요한 attribute에 대해 위쪽에 있는걸 쓰라는 말이겠군.

---

Accuracy

Steven’s Power Law

Channel accuracy by data type!

Crowdsourced Results

• Position
  • Bar chart within group
  • Stack chart at bottom
  • Bar chart between groups
  • Stack chart both top
  • Stack chart within group
• Angle
  • Pie Chart
• Areas (Circular)
• Areas (Rectangular)

→ 어느정도 기존 연구와 동일한 순서임

---

Discriminability

Line width → only a few

• Resolution을 1~2px 단위로 설정할 수 없다.

Separability, Integrality

• Position + Hue: Fully Separable
• Red + Green: Integral Hue

---

Preattentive Processing (Popout)

• Cognitive operations done preattentively, without focused attention

→ popout effect, segmentation effect

• Parallel processing on many individual channels
  • speed independent on # of distractor
  • speed depends on channel and amount of channel difference
• Serial search depends on # of distractor → Almost all combination

---

Gestalt Psychology

the whole is different from the sum of its parts.

→ How it affects?

1. Containment
2. Connection
3. Proximity (spatial region이 비슷하면 하나의 group)
4. Similarity (같은 shape, color면 하나의 group)

---

Graphic Design

Contrast

guide the users’ attention to key element by contrast

→ preattentive

Repetition

create visual consistency

Alignment

important, because position is most effective channel!

→ Creates visual connection

Weber’s Law

• Perceptual system mostly operates with relative judgements, not absolute.

→ Perceived change is propitional to relative stimuli.

원래 길이가 길면 그 차이를 더 크게 줘야한다. (원 자극이 컸다면 차이를 just noticeable하게 해주기 위해 더 큰 변화를 줘야함)

\[\frac{\delta I}{I} = K\]

I is intensity.

---

Lightness Constancy

Luminance는 perceived lightness와 무관하다.

→ 검은 종이가 밖에 있는 것과 흰 종이가 실내에 있는것.

1. Photoreceptor adaptation: 막대세포와 원추세포가 알아서 적응함
2. Relative coding: 주변 환경에 비해 얼마나 밝은지 해석 (비율 고려)
3. Perceptual Consistency: 벽이 흰색인걸 알고 있다면 그 기억을 이용