Oplægget blev holdt ved InfinIT-arrangementet Temadag om Interaktionsdesign, der blev afholdt den 20. juni 2013. Læs mere om arrangementet her: http://www.infinit.dk/dk/hvad_kan_vi_goere_for_dig/viden/reportager/styr_din_mobiltelefon_med_et_nik.htm

1. Displayless
Interac/on
Lars
Bo
Larsen
Mai
Baunstrup,
Søren
Staal
Jensen
Infinit
Temadag,
AAU
20
juni
2013

2. Outline
• Introduc/on
of
the
Displayless
Interac/on
(DI)
project
• Presenta/on
of
AAU
contribu/on:
– “Using
sound
content
to
direct
user’s
aOen/on”
– Experiments
– (preliminary)
Results
2

3. Display-less Interaction "
Outline of the Display-Less Interaction Project"
– To investigate interaction design for devices without
displays"
– Partners: Bang & Olufsen, Radiometer, AAU"
– Infinit “mini project”, running January 1 – June 30
2013"
– Close integration with PDP thesis students
(Engineering Psychology)"
"
3

4. Display-less Interaction "
Our Focus"
– Using the auditory modality "
– To investigate if designed sounds can guide the
user’s attention onto certain areas of a user interface"
– To achieve this with sounds that are intuitive, which
do not have to be learned (earcons) or to be based on
metaphors or cultural references (auditory icons)"
4

5. Background"
• Pratt (1930) found that the pitch of tones were related to
their perceived positions"
5

6. Background"
• Walker and Ehrenstein (2000) tested 12 sounds that
either increased or decreased in pitch"
6
• Results: Test-persons
matched an increase
and the decrease in
pitch with ”high” and
”low”, at a 96% accuracy
level"
"

7. Theories"
• R. Walker (1987) ”The effects of culture, environment,
age, and musical training on choices of visual metaphors
for sound”"
• 828 test-subjects were tested to see how they matched
acoustic features (frequency, waveform, amplitude, and
duration) to visual metaphors on a response sheet. "
"
• Left-Right was generally matched with the duration of a
sound"
7

8. Our Experiment"
Inspired by this, we decided to design a number of sounds with the
goal of directing the user’s attention in a particular direction on a plain
surface."
"
Inspired by the fact that earcons, designed as musical sounds are
considered more attractive to listen to than other types, we chose a
format of short piano sounds or tunes."
"
12 sounds were designed with the help of a musician:"
– 2 pitch change sounds (1 increasing, 1 decreasing)"
– 2 high / medium high pitch tones"
– 2 low / medium low tones"
– 2 sounds with increasing/decreasing beat"
– 4 hybrid sounds (with beat and pitch change)"
• Duration is 1.1-1.2 seconds"
• Tones are roughly within one octave"
8

9. 9
Our
speaker
device
consists
of
16
individual
speakers
(built
by
B&O)
set
in
a
frame
The
speakers
are
connected
in
4
groups
with
4
speakers
in
each
corner
This
allows
to
define
5
channels:
Le^,
Right,
Up,
Down
and
All
In
the
experiment,
the
speakers
are
hidden
to
the
test
person

10. Experimental design"
We designed two experiments:"
1. Half the sounds are played by all speakers, (ie the
sound is emitted uniformly from the surface of the
device). The other half were played in either Right,
Left, Up, Down sections as a control condition"
– 30 test persons each listened to 192 sounds"
2. All sounds are played by all speakers"
– 16 test persons each listened to 96 sounds"
• Only results from experiment 2 is reported here."
10

11. Experiment setup"
• The test person
used a laser pointer
to indicate where
she experienced
the sound "
• The facilitater
visually determined
the position in a
4x4 grid"
• The playback and
recording of
positions were
controlled by a
matlab program"
11

12. Experiment 1"
• Data was obtained via a matlab program,
where the test-persons response was
recorded via a push on a button, in
accordance to a 4x4 grid"
13

13. Experiment 1 Results"
• 30 test-persons participated"
• The general results are that the acoustics
were dominant in accordance to how the
users gave their response"
• The sounds only had an influence within
these areas. "
14

14. The decrease in pitch sound"
15

15. One of the diagonal ”up” hybrid sounds"
16

16. Experiment 1 Results"
• Notable answers from the questionnaires "
17

17. Experiment 1 Results"
• Other interesting data"
Feedback from an explorative question What many test-persons mentioned "
"
18

18. Results:"
Experiment 1:"
– showed that the acoustics was the dominant factor – ie when
sounds are played from different positions, users will hear this
and it overrides the contents"
– Analysis of interdependencies, etc are ongoing."
Experiment 2:"
– The purpose of Experiment 2 is to see what happens when the
factor of acoustics is eliminated"
– Preliminary results show clear effects in some cases and not so
clear in others, analysis is still ongoing"
19

19. Experiment 2 results"
Pitch and beat change"
"
"
"
" Decrease " "Increase"
"
"
Decrease "Increase"
20
Pitch
Beat"

20. Experiment 2 results"
• Static Pitch"
21
High
Medium
high
Medium
Low
Low

21. Experiment 2 results"
• Hybrid Sounds"
Pitch decrease + beat up Pitch decrease + beat down"
"
Pitch increase + speed up Pitch increase + beat down"
22

22. Experiment 2 results"
• Answers from the questionnaires regarding whether the
sounds could guide the test-persons up or down"
23

23. Experiment 2 results"
• Answers from the questionnaires regarding whether the
sounds could guide the test-persons left, right or into a corner"
24

24. Statistical tests (Chi-Square)"
Increase in pitch (”up sound”)"
"
"
"
Yes,
significant
Decrease in pitch (”down sound”)"
"
"
Yes,
significant
25

25. High static pitch"
"
"
"
"
"
" " " " " " " " " " "Yes,
significant
Medium high static pitch"
"
•
Yes,
significant
26
Statistical tests (Chi-Square)"

26. Statistical tests (Chi-Square)"
Increase in beat (”right sound”)"
"
"
"
"
Not
significant
Decrease in beat (”left sound”)"
"
Not
significant
27

27. Medium low static pitch"
"
"
Low static pitch"
28
Statistical tests (Chi-Square)"

28. Summary"
• Experiment 2 showed that the pitch level and changes
were the dominant factor"
• The pitch up and down sounds worked as expected"
• The static pitch worked for the high tones, but only partly
for low"
• The beat change – ie left and right sounds did not work"
• The diagonal and right-left sounds did not work as
intended"
• The diagonal and right-left sounds worked as up and
down sounds, in accordance to their pitch change"
29

29. Application"
So, how do we apply this result for
displayless Interaction?"
– Our goal was to show that by utilising
specific designed sound patterns, a
sound source location can be
simulated under certain circumstances,
without any added mental load on the
user"
– This has implications for displayless
devices, where e.g. a blinking LED can
be replaced by a “down” sound to
direct the users attention downwards"
– Leading to better interaction (probably)
better aesthetics and cheaper
manufacturing costs"
30

30. Conclusions
A
first
step
towards
a
design
guide
for
“simulated
3D”
sounds
have
been
taken
• Preliminary
results
show
some
clear
indica/ons
for
up-­‐down,
however
le^-­‐right
s/ll
remains
unsolved
• There
are
interdependencies
when
the
physical
sound
source
can
be
located
• We
have
deliberately
avoided
metaphorical
sounds
and
chosen
very
generic
paOerns
to
leave
designers
as
much
freedom
as
possible
• Sta/s/cal
analysis
is
s/ll
ongoing
• Results
are
applicable
and
will
reduce
manufacturing
costs
and/or
increase
usability
31