Stanford hci groupFeb 9, 2009 Björn Hartmann bjoern@cs.stanford.edu Understanding & Modeling...

stanford hci group Feb 9, 2009

Björn Hartmannbjoern@cs.stanford.edu

Understanding & ModelingInput Devices

2

Questions for today

1. How do common input devices work?

2. How can we think about the larger space of all possible input devices?

3. Can we predict human input performance?

Next class: What about uncommon input devices (music controllers, multitouch, …)?

3

Today’s lecture in graph form

time

Level of abstraction

concretedetails

abstractmodels Functional

Dissection of Mouse & Keyboard

Design Space of

Input Devices Modeling

Human Performance

I spilled coffee on my keyboard.

Now 25% of the keys don’t work anymore.

But some of the defective keys are nowhere near the spill.

What’s going on?

5

Key cap

Top conductive layer

Bottom conductive layer

Separating layer(with hole)

Key cap

Top conductive layer

Bottom conductive layer

Separating layer(with hole)

9

10

Row/Column Scanning

Q W E R T

A S D F G

Z X C V B

R1

R2

R3

R4

C1 C2 C3 C4 C5

9 lines

20 keys

Mouse. Engelbart and English ~1964Source: Card, Stu. Lecture on Human Information Interaction. Stanford, 2007.

A Layered Framework

12From: Hartmann, Follmer, Klemmer: Input Devices are like Onions

13

14

15

Right button

Left buttonEncoder wheel for scrolling

16

IR emitter IR detectorslotted wheel(between emitter & detector)

17

Sensing: Rotary Encoder

High

18

Sensing: Fwd Rotation

Low

19

Sensing: Backwd Rotation

Low Oops!

20

Solution: Use two out-of-phase detectors

High

High

21

Sensing: Rotary Encoder

Low

High

22

Sensing: Rotary Encoder

High

Low

Coding:HH-> LH: dx = 1HH-> HL: dx = -1

23

Transformation

cxt = max(0, min( sw, cxt-1+dx*cd ))

cyt = …

cxt: cursor x position in screen coordinates at time tdx: mouse x movement delta in mouse coordinatessw: screen widthcd: control-display ratio

24

Device Abstraction Click, DoubleClick, MouseUp,

MouseDown, MouseMove …

25

What about optical mice?

Source: http://spritesmods.com/?art=mouseeye

26

bbbbbbbbbb

Source: http://spritesmods.com/?art=mouseeye

27

Trackball, Trackpad

28

Trackpoint Indirect, force sensing, velocity

control Nonlinear transfer function

Force

Velo

city

(cc) Image by flickr user tsaiid

29

Joysticks

A design space of input devices…

Card, S. K., Mackinlay, J. D., and Robertson, G. G. 1991. A morphological analysis of the design space of input devices. ACM TOIS 9, 2 (Apr. 1991), 99-122.

Implicit Assumptions: Desktop-centric computing

33

Which device is fastest? For what task? Pointing.

Combination of two factors: Bandwidth of human muscle group

(upper limit) Bandwidth of device itself

Bandwidth of Human Muscle Groups

Source: Card, Stu. Lecture on Human Information Interaction. Stanford, 2007.

Fitts’ Law Time Tpos to move the hand to

target size S which is distance D away is given by: Tpos = a + b log2 (2D/S)

Time to move the hand depends only on the relative precision required

Source: Landay, James. “Human Abilities”. CS160 UC Berkeley.

Mouse vs. Headmouse

Source: Card, Stu. Lecture on Human Information Interaction. Stanford, 2007.

Headmouse: No chance to win

Source: Card, Stu. Lecture on Human Information Interaction. Stanford, 2007.

Fitts’ Law in Windows & Mac OS

Windows 95: Missed by a pixelWindows XP: Good to the last drop

The Apple menu in Mac OS X v10.4 Tiger.

Source: Jensen Harris, An Office User Interface Blog : Giving You Fitts. Microsoft, 2007; Apple

Fitts’ Law in Microsoft Office 2007

        Larger, labeled controls can be clicked more quickly

Mini Toolbar: Close to the cursor

                Magic Corner: Office Button in the upper-left corner

Source: Jensen Harris, An Office User Interface Blog : Giving You Fitts. Microsoft, 2007.

stanford hci group Feb 9, 2009

http://bjoern.org