Basics of Stimuli

1
Basics of Stimuli

• John H. Krantz
• Hanover College

2
Outline

• The nature of the beast
• Basic Perceptual Issues
• Spatial
• Temporal
• Color
• Luminance

3
The General Beast
4
Screen Mosaic Triad Arrangement
5
Screen Mosaic Color Strip Arrangement
6
Interlaced Projection
7
Progressive Projection
8
The Electron Beam
9
Seeing the Flicker
If cant get movie by clicking on the picture,
click here.
10
Spatial Issues

• Spatial Inhomogeneity
• Aliasing
• Edge Effects and Contrast

11
Spatial Inhomogeneity

• The same output value leads to different
  luminance outputs at different screen locations.
• Thus, you get different outputs at different
  screen locations.
• In general, draw to center if critical

12
CRT Change from Minimum
13
How to Handle Spatial Inhomogeneity

• Keep stimuli in central region

14
Aliasing

• Technical definition
• When an image contains frequencies beyond the
  range of the sampling matrix, the wrap-around and
  occur as lower frequencies, distorting the image

15
Aliasing

• Description of aliasing
• Generally try to draw and image that has too fine
  of detail or sharp edges
• Causes jaggies
• And pixel artifacts
• That is you can notice the pixels

16
Aliasing and Pixel Effects
17
Craik-Cornsweet
18
Craik-Cornsweet Described
The figure above is an exaggerated map indicating
the light levels across the image on the previous
slide. Note how the center and edges have
identical luminance. That can be seen by sitting
far enough away from the screen
19
Minimal Contours
20
Minimal Contours Described
There are two circles below. Both circles have
the same luminance (intensity level) at the
center. Click on your mouse and This one changes
abruptly watch as the edges are blurred to
the level at the center. and the circle
disappears.
21
Anti-aliasing

• Solution to aliasing
• Bandlimiting or anti-aliasing
• Essentially blur the edges or image so fine
  detail is lost
• In a lot of Software
• Ideally use a gaussian filter
• This is the shape of CRT electron beam which is
  why dots on screen never show up
• Infante (1985) Silverstein et al. (1990)
• Many graphics software and even some languages
  anti-alias (Java2D)

22
Temporal Issues

• Motion reproduction
• Flicker
• Stimulus Timing

23
Motion Reproduction

• Frame rates of monitors far exceed what is needed
  for movement update
• NTSC video 30 updates per second
• Movies 24 updates per second
• Most web video 10 to 15 frames/sec

24
Computer Video

• Most Update rates are 10-15 frames/sec
• 5 fps 10 fps 15 fps

Click on pictures to start each of the movies.
If that does not work click here for the left
hand movie, click here for the middle, and here
for the right.
25
Flicker

• Critical threshold of flicker rates is about 60
  Hz in the fovea
• But gets higher for larger stimuli
• Recommended flicker rates between 66 Hz and 120
  Hz (Bridgeman, 1998)
• Most monitors are adequate in this value
• Notice difference between flicker and frame
  update rate.

26
Stimulus Duration

• Pixels are not on continuously during a frame
• In part this is necessary for clean motion
• Typical CRT phosphors last about 4 msec.
  (Bridgeman, 1998)
• On LCD and other technologies, persistence is
  longer
• Makes motion less clean but flicker less
  noticeable

27
Differences in Persistence
Click on pictures to start movies. If that does
not work, click here for the crt and click here
for the lcd.
28
Determining Stimulus Duration

• Possible errors
• Not knowing which frame the stimulus is first
  presented on
• Assuming the stimulus is visible from the
  beginning of the display
• Stimuli are not visible at beginning of frame but
  some time later
• Delay depends upon where on frame stimulus is and
  frame rate
• This is a constant error

29
Computer Video Architecture (Win95)
Display
Win32 Applications
GDI DDI DCI DDI 3-D DDI
GDI Video OpenGL 3-D Games
30
Effects on Timing

• If no more than one frame delay
• Additional error of up to 17msec added to all
  conditions
• No longer msec timing but really 60 Hz timing
• If frame delay is larger effects are worse

31
Example Results from Myors (1999)
32
Luminance Issues

• Gamma

33
Gamma
34
Gamma Correction

• Typically well described by LL0a(V-V0)g
• g is typically near 2.3 L, L0, a, g are affected
  by contrast and brightness settings
• Cant be done on web
• Make sure stimuli are not sensitive to
  distortions in gray scale
• High contrast
• Photos may be distorted slightly and impact
  results (Ruppertsberg et al., 2001)
• Dont require fine distinctions

35
Chromatic Issues

• The Trichromatic Theory
• Color Picker
• The Color Gamut
• Glare
• Color Blindness

36
The CIE Color System

• A set of Equations that
• allow predictions of
• matching.
• Used in photo printing,
• TV and film.
• Wyszecki Stiles(1967)

37
The Color Gamut

• The range of colors reproducible by any system

38
The Effect of Illuminance on Gamut
In Dark With Lights on With Lights on and
30 Luminance Level
39
Illuminance

• Agostini Bruno (1996)
• Accuracy of the perceived stimulus is affected by
  the amount of illuminance.
• D. F. Neri (1990)
• Combination of light and gamut changed
  chromaticity
• Request close windows and turn off unnecessary
  light
• Do not use subtle color differences

40
Color Blindness
Description
Click on picture to start movie. If that does
not work, click here.
41
Color Blindness
Click on picture to start movie. If that does
not work, click here.
42
Color Blindness
Click on picture to start movie. If that does
not work, click here.
43
Color Blindness
Click on picture to start movie. If that does
not work, click here.
44
How to handle

• Screening questions?
• Large sample and random assignment
• Pre-test on color blind

45
Summary

• Spatial
• Aliasing
• Edges
• Temporal
• Frame per second in video
• Luminance
• Gamma
• Color
• Glare
• Color deficiencies