Mixing Interaction, Sonification, Rendering and Design - The art of creating sonic interactions
-
Upload
eoin-brazil -
Category
Education
-
view
1.281 -
download
0
description
Transcript of Mixing Interaction, Sonification, Rendering and Design - The art of creating sonic interactions
Mixing Interaction, Sonification, Rendering and Design - The art of creating sonic interactions
Eoin Brazil
EG_IT 22 & 23 October, Verona. Italy
Sonification
Using Sound to Effectively Convey Information
Ben Cohen, co-founder of Ben & Jerry’s Ice-cream Company
Talking against the size and cost of the nuclear stockpile in the United States.
Cartoonification
Maisy drinking with a straw
(From “Happy Birthday, Maisy” by Lucy Cousins, Candlewick Press, 1998.)
Chapter 9. High-level models: bouncing, breaking, rolling, crumpling, pouring 203
Figure 9.17: Maisy drinking with a straw (From “Happy Birthday, Maisy” byLucy Cousins, Candlewick Press, 1998.)
audio-cartoon of a filling process.As an example for a more abstract use of our sound objects, serves a son-
ification of a scene from children’s book. The combination of the streaming-object with a comb-filter could be interactively controlled simultaneously witha simple adjustable cardboard figure in the book: an “aristocratic mouse”,drinking lemonade with a straw! In the book, the reader can regulate the fluidlevel of the mouse’s drinking glass (see Figure 9.17).
That control might be coupled with the delay time (i.e. base frequency)of the comb filter (see Figure 9.18). Together with typical envelopes for astarting/finishing streaming-action an additional inspiring acoustic feedback isgiven. Of course, drinking with a straw generally does not generate a soundclose to this one, if any remarkable noise at all. Nevertheless, the free associ-ation of this moving liquid-like sound, with that cartoon scene, seems natural.Asked to spontaneously sonify this scene with their voice, many people wouldprobably produce similar noises, although it does not really sound like having-a-cocktail (hopefully).
Area-1 Area-2 Area-3 Area-4 Area-5 Area-6 Area-7 Area-8
0.333 1 0.333 0.692 0.176 0.138 0.354 0.058
0.554 1 0.554 1.232 0.074 0.198 0.358 0.065
1.028 1 1.028 0.576 0.070 0.077 2.401 0.162
0.217 1 0.217 0.885 0.259 1.087 0.739 1.005
0.507 1 0.507 0.862 0.198 0.309 0.578 1.107
0.570 1 0.570 0.866 0.301 0.434 0.526 0.674
0.554 1 0.554 1.232 0.074 0.198 0.358 0.065
0.530 1 0.530 0.042 0.852 0.250 0.053 0.185
0.497 1 0.497 1.590 1.365 1.175 1.278 0.101
0.167 1 0.167 1.261 0.031 0.658 0.306 1.289
0
0.4
0.8
1.3
1.7
2.1
2.5
Area-1 Area-2 Area-3 Area-4Area-5 Area-6 Area-7 Area-8
An Example of Auditory Graphing
Complex datasets - Adding sound can elaborate streams and temporality of the data
Area-1 Area-2 Area-3 Area-4 Area-5 Area-6 Area-7 Area-8
0.333 1 0.333 0.692 0.176 0.138 0.354 0.058
0.554 1 0.554 1.232 0.074 0.198 0.358 0.065
1.028 1 1.028 0.576 0.070 0.077 2.401 0.162
0.217 1 0.217 0.885 0.259 1.087 0.739 1.005
0.507 1 0.507 0.862 0.198 0.309 0.578 1.107
0.570 1 0.570 0.866 0.301 0.434 0.526 0.674
0.554 1 0.554 1.232 0.074 0.198 0.358 0.065
0.530 1 0.530 0.042 0.852 0.250 0.053 0.185
0.497 1 0.497 1.590 1.365 1.175 1.278 0.101
0.167 1 0.167 1.261 0.031 0.658 0.306 1.289
0
0.4
0.8
1.3
1.7
2.1
2.5
Area-1 Area-2 Area-3 Area-4Area-5 Area-6 Area-7 Area-8
An Example of Auditory Graphing
Complex datasets - Adding sound can elaborate streams and temporality of the data
Sonic Interaction Design
• Sound and interaction are both process oriented
• Interactive systems are complex - sound is able to convey complex information very quickly
• Interactive systems are “disappearing”, sound can tell us about their presence and their status
• We have learned to understand complex soundscapes -> ubiquitous, visualisation, and many areas of computing may profit from the use of sound
SHAPE ExhibitionRe-Tracing the Past
Interactive Sound Object Models for Exploration & Education
SolidsFractureModel
Braking
Breaking Sliding
Footsteps on gravel
Can crushing
Falling coin
Rolling / breaking
wheel
Squeaking door
Paper crumpling
Struck string Struck bar Glass
cleaningBowed string
Rubbed glass
Fracturing
Walking
Crumpling
Crushing Bouncing
Hitting Rolling Rubbing
SqueakingDropping
ImpactModel
FrictionModel
Fluid flowingModel
Flowing Burbing
Splashing
BubbleModel
Dripping
FillingPouring
TurbulenceModel
Whoosing
Sucking
Burning Exploding Popping
Vacuum cleaner
Explosion Model
Liquids GasesLow LevelModels
Basic Eventsand Textures
DerivedProcesses
SimulationExamples
SOb-ifying the Exhibition
Impact Models for Hitting, Rolling, or Crumbling Textures / Basic Events
Virtual Touch
MachineInteracting with one of the four `mysterious’ objects
Using a multimodal (Audio-Haptic-Visual) interface to explore its materiality and size
Real-Time Interaction
Virtual Touch
MachineInteracting with one of the four `mysterious’ objects
Using a multimodal (Audio-Haptic-Visual) interface to explore its materiality and size
Real-Time Interaction
SOb-ifying the Exhibition
SOb - Sounding Object Project
Interactive Sound Object Models for Exploration & Education
Museum Installations - Lessons & Considerations
• Enhance interpretation
• Fun yet interactive, spark
• Active not passive environment / interfaces
• Design activities / consultations
✦ observational studies
✦ consult docents /experts
✦ understand physical/spatial constraints
Soundscapes & Train Stations
J. Tardieu, P. Susini, F. Poisson, P. Lazareff, and S. McAdamsFrance, 2008
Soundscapes & Train Stations
J. Tardieu, P. Susini, F. Poisson, P. Lazareff, and S. McAdamsFrance, 2008
BouncingBalls
Frequency / Amplitude
Airplane Synth
Mirror Triggers
Connecting Strangers at a
Train Station
L. Pellarin, N. Bottcher, J. M. Olsen, O. Gregersen, S. Serafin, and M. GuglielmiDenmark, 2005
Shannon Portal - Auditory Display for
Public Spaces
Interactive Sounds for Communication & Entertainment
Ping Photon Drive
Wind Shepard Tones
Shannon Portal - Auditory Display for
Public Spaces
Interactive Sounds for Communication & Entertainment
Ping Photon Drive
Wind Shepard Tones
Shannon Portal - Auditory Display for
Public Spaces
Interactive Sounds for Communication & Entertainment
Ping Photon Drive
Wind Shepard Tones
Shannon Portal - Auditory Display for
Public Spaces
Interactive Sounds for Communication & Entertainment
Ping Photon Drive
Wind Shepard Tones
Shannon Portal - Auditory Display for
Public Spaces
Interactive Sounds for Communication & Entertainment
Ping Photon Drive
Wind Shepard Tones
Analysis of the Sound & Space
Proceedings of the 15th International Conference on Auditory Display, Copenhagen, Denmark, May 18-22, 2009
ICAD09-3
3.1. First design: Ping
Our first display used an 80 ms decaying pulse from two
sine wave oscillators 4 Hz apart, pitched from 200 Hz to 800
Hz. The rate of pulses ranged from 0.3 to 5 Hz. The output was
amplitude panned left-right. In essence, one could describe this
simple sound design as “a machine that goes ping”. We tried the
following mapping:
• Amount of movement – Pulse rate
• Left-Right – Panning
• Front-Back - Pitch
Participants found this display to be clear but too intrusive
(“sounds like a truck reversing”). Part of the intrusiveness in the
sound design is due to the sharp onset, which is strongly
attracting attention. The pitch used was in the middle of the
background soundscape spectrum, hence would require to be
loud to avoid masking. All participants found the mapping easy
to understand.
3.2. Second design: PhotonDrive
Our second display used a pulse with approximately 0.3 s
attack and 0.3 s release from two sawtooth oscillators tuned 2%
apart, pitched from 50 Hz to 650 Hz. The output was amplitude
panned. Our nickname for this sound design was PhotonDrive.
The mapping used in our tests was the same as in the previous
trial.
The softer onset and richer spectrum resulted in participants
feeling that the sound design was less intrusive. Due to the
richer spectrum, the sound could be played at lower loudness
level, but was still found to be masked at times.
3.3. Third design: Wind
In our third display we tried a less intrusive sound, a wind-
like sound that was based on subtractive synthesis using white
noise sources that were bandpass filtered with a certain amount
of jitter in the center-frequencies to make it sound more
realistic. The average center frequency was mapped to the
amount of movement.
This sound was noted as not being intrusive as there was no
distinct onset, but due to its naturalness, basically being a
parametrically controlled auditory icon, and only two mappings
(activity – frequency, left-right panning), also this display was
discarded.
3.4. Fourth design: Shepard tones
In our fourth and final display we used a slightly more
radical approach. This sound design was based on the Shepard
tone illusion (Shepard, 1964). This illusion is perceived as
eternally rising or falling pitch and is due to the continuous
frequency change and mix of ten or more partials. We carefully
adjusted the partials to a quite unnatural timbre (bell-like) to try
to avoid masking effects in relation to the surrounding
soundscape.
• Amount of movement – loudness
• Left-Right – Panning
• Front-Back – Pitch direction
Participants found this display to be positively intriguing, as
it did not sound like anything they had heard before. Due a soft
onset, as we mapped amount of movement to loudness, it was
noted as both being pleasing as well as attention catching. A
participant or group of participants walking past and
accidentally being tracked and sonified immediately understood
that it was their movement that caused the sound. This finding
is similar to Todd Winkler’s installation Light around the edges
[10].
We finally moved the installation to Shannon airport where
it was installed for 3 weeks. We carried out both observations
and informal interviews during this period. Both passengers and
staff at the airport found the sound design contributing to their
experience and being non-intrusive, and our previous findings
confirmed. Our video analysis of people using the system on
site also confirmed that the auditory display served to catch
people’s attention when walking past the system.
In Figure 4, below, you can see an example sonogram of the
soundscape of Shannon Airport with the Shannon Portal
installed. Within the white square, you see the partials of the
auditory display increasing in frequency. The rest of the
sonogram is the background noise. Even in this visual
representation, the sonification stands out against the
background. The psychoacoustic features we are exploiting are
‘common fate’ and ‘good continuation’ [11].
Figure 4: Shannon Portal soundscape sonogram
4. DISCUSSION
There are a number of issues involved in this design. First
of all, it was important to create an auditory display that would
be somewhat subtle but always audible, but not annoying for the
staff working in the area. It also had to be immediately
informative, i.e. affordances [12-14] for immediate discovery
by casual users. The auditory display also had to be useful, i.e.
to actively provide users with information that increased their
feeling of engagement with the system.
Our overall design ambition was to develop a ubiquitous
computer system for a public setting. Instead of using
traditional metaphors, we aimed for a high degree if virtuality
[15], i.e. that the conceptual structure of the system would be
immediately obvious to the users instead of being weighed
down by traditional user interface clichés or forcing users to
read a manual before trying to use the system. This is always a
problem in computing for public environments, as most users
would never have encountered a system like this before (at
least, not the first time) and within a few seconds the system has
to communicate to potential users that it can interacted with and
how to proceed.
Throughout the design iterations, the first three designs had
some problems to fulfill the requirements, but the fourth and
final iteration was found to be a satisfactory solution to the
problem. We believe that the important finding here is to
facilitate multiple design iterations with evaluation, to help
Linking ubiquitious auditory displays design to evaluation in
public environments
Evaluate
Brainstorms, Concept Generation
Sonic Mood Boards
Sketching sound using clips (mood boards or
onomatopoeia)
Bodystorming / Roleplay the
Inteaction
Video Prototyping
Evaluate
Prototype using PD / Max/MSP or similar
Evaluate
Code interaction and sound object model
Evaluate
Auditory Public SpacesAuditory
Display Design Process for
Public Spaces
Proceedings of the 15th International Conference on Auditory Display, Copenhagen, Denmark, May 18-22, 2009
ICAD09-1
The Shannon Portal: Designing an Auditory Display for Casual Users in a
Public Environment
Author 1
Address
Address email@host
Author 2
Address
Address email@host
ABSTRACT
We developed an installation for a public environment and
casual users where auditory display was a significant element to
facilitate user interaction. We used an iterative design process,
starting from simple onomatopoeic representations, to complex
sound object models in Pure Data. The system was evaluated at
each stage, from the lab to the final public setting. The problems
addressed covered the representations of left-right, up-down,
and the amount of movement by the user or groups of users. In
addition to this, it was important that the auditory display would
attract attention when users were within control range of the
system, i.e. an affordance that invited and allowed users to
discover functionality.
1. INTRODUCTION
In our Shared Worlds project we explored designing
ubiquitous computing for public environments. Part of this
exploration was to design, build and evaluate installations in
public environments. The work reported in this paper is about
the Shannon Portal, a complex interactive installation in
Shannon Airport in the West of Ireland. The overall design is
described in [1], while in this paper we focus on the design of
the auditory display element of the installation. One of the
issues that we wanted to explore in the project was the use of
large ambient displays and public interaction with such
displays. As the installation at Shannon required an interactive
image gallery to be displayed, we designed a purpose-built
back-projection system. An overhead camera tracked users’
movements in front of the display, allowing the users to move a
virtual magnifying glass across the image gallery on screen, by
moving their body in front of the display. See Figure 1 and 2.
Over several iterations we developed hardware and software
for the display and explored different mappings of users’
actions to movement of the virtual magnifying glass on the
projection screen. For reasons outside the scope of this paper, it
was decided that it was not the users’ location that would
control the movement, but the users’ movement. The rationale
for this was that as there was only one virtual magnifying glass
but there could be multiple simultaneous users, this mapping
would encourage collaboration and engagement.
While testing and evaluating the system in the lobby outside
our lab in the Interaction Design Centre at the University of
Limerick, we noted that users who occasionally passed by the
display were not aware of that the display was interactive.
Furthermore, due to the nature of the mapping, some users
found it difficult to understand that it was their movement, not
their location, that controlled the virtual magnifying glass. To
remedy this, we decided to design an auditory display element
for the system.
Figure 1: Shannon Portal installation
Figure 2: Back-projection, loudspeakers and video
camera
1.1. The challenge of auditory display in public
environments
Is has been noted in numerous papers (e.g. [2-4]) that
auditory display can be annoying. A location such as the transit
hall of an airport has an inherently high background noise level,
at peak-time often over 85 dB. In the transit hall in Shannon
airport, the main contributors to the soundscape are people
walking, talking, handling luggage, eating and drinking, mobile
phone ring tones, tills in the Duty Free shop, all the activity in
the Sheridan Bar and bursts of announcements about flights
Left-Right Panning = Amplitude Panning
Front-B
ack =
Pitch
Direction
Movement within space = Loudness
1 2
3
Mappings used for fourth prototype (Shepard Tones)
Public Spaces - Lessons & Considerations
• Engaging (virtuality)
• Visitors / Working within the space
• Privacy / Intrusion
• Masking and soundscape issues (peak vs non-peak times)
• Design activities / consultations
✦ observational studies
✦ interviews
✦ understanding the physical /cultural / spatial constraints
Hanging Gardens - Auditory Installation
Collaboration with Jurgen Simpson
Linked two gallery spaces in Limerick and Belfast
Explore circadian rhythms of the spaces
Motion sensors and time of day, site-specific
Artistic Installations - Lessons & Considerations
• Interpretation vs artistic
• Engaging
• Group / Individual
• Consider temporal constraints
• Design activities / consultations
✦ aesthetic issues✦ distance /
repairability✦ understand
physical/spatial constraints
HCI Design and Interactive Sonification for Fingers and Ears
Action Sound Function
No touch N/A
Touch area outside button N/A
Enter button area Tick
Move finger on button Friction sound
Exit button area Tack
Lift finger off button Tock Select / Activate function
Where and How to start designing with Sound
• How to select and classify suitable sounds for a particular interaction design ?
• TaDa - select sounds to communicate the dimensions and directions with relation to their actions and to that of the system e.g. ears-lead-eyes design pattern
• Sounds linked to physical objects as dynamic entities with manipulation / interaction
• Multidimensional information carrier - size, shape, material, distance, speed, and emotional expression
Selection of Approaches Depending on Question
• Evaluating Perceptual Qualities of Single Auditory Stimuli
• Identification Tasks, Context-Based Ratings, Attribute Ratings
• Evaluating Relations Among Auditory Stimuli
• Discrimination Trials, Similarity Ratings, Sorting Tasks, MDS, CA, PCA, Pathfinder Analysis
• Evaluation of an application or an interface
• Surveys, Verbal Protocols, Heuristics, Task lists, etc.
Sound Creation
Sound Analysis
Context and Auditory Display Definition
ConsiderationsActivity, Sound Like,
Vocalisations / Onomatopoeia
Find Closest Match in F/X Library / Use
Foley
Analyse Sound andConsider Action/s and
Objects making it
Create the Sound and Test - Does it sound
right?
Musical Sounds
Everyday Sounds
ConsiderationsActivity, Mapping,
Instruments
Explore Existing Earcon Hierarchies /
Create New Hierarchy
Create Earcon Family and Test - Does it
sound right?
Evaluate Scalings / Mappings
Elicit Descriptors & Constructs
Rating of Constructs & Descriptor Categorisation
Personal ConstructsAssociated RatingsDescriptor Categories
Causal Uncertainty Measures Structuring Of Constructs
Definition Of AttributesConstruction Of Scales
Validation Of Scales
Category Refinement
Evaluation
Results
Attribute Scales
Groups of Constructs
Preliminary Results
Confused Sounds
Inconsistent Attributes or MappingsConfused Sounds
Inconsistent Attributes or MappingsConfused Sounds
Auditory Characterisation of Story / Scene / Account
TaDa Analysis Sonic Map Creation for Account / Scene / Story
Structuring Of ConceptsHearsay Analysis
Groups of Concepts
Evaluation
ResultsInconsistent Attributes or MappingsConfused Sounds
Mismatched Sounds
Narrative Sound Artefact Creation
1
2 2
22
2
3&4
3&4
5
7
12 13
14
15
16
17
6
8 8
11
17
9
10
Sound Creation
Sound Analysis
Review of existing approaches,
methods, and techniques.
Select Listening Panel, Sounds
& Attributes to explore
Play audio stimuli (~100-200)
Sta
ge
1 -
Dev
elo
p c
ons
ensu
s
des
crip
tive
lang
uag
e
Determine attribute scales,
descriptors, key audio groups
Sta
ge
2 -
Par
ticip
ant
per
form
ance
dat
a
Gather end
user
preference
Gather global
scores from
naive
participants
Sta
ge
3 -
Map
pin
g
dat
a to
pre
fere
nces
Map preference / performance
data using statistical methods
Select Listeners, Sounds &
Attributes to explore
Play audio stimuli (~100-200)
Stag
e 1 - Develo
p ind
ividual
descrip
tive languag
e Determine attribute scales,
descriptors, key audio groups
Stag
e 2 - Particip
ant
ranking / rating
of d
ata
Participant ranks, rates, or
groups data
Stag
e 3 - Map
ping
of d
ata
Mapping of data using
statistical methods
Group or Individual based approaches
Functional ArtefactsLinking design to evaluation
including basic design methods
Sonic Interaction Design Process for
Functional Artefacts
Interaction MaterialsAnalyse,Describe
Interaction Gestalt
Experience in Context
Interaction Artefact
Shape, Combine
Evaluate
Functional Artefacts CLOSED project, 2008, Ircam, Verona, Berlin
Rolling
Clapping 1
Clapping 2Examples:
Functional ArtefactsLinking design to evaluation
including basic design methods
Sonic Interaction Design Process for
Functional Artefacts
Interaction MaterialsAnalyse,Describe
Interaction Gestalt
Experience in Context
Interaction Artefact
Shape, Combine
Evaluate
Functional Artefacts CLOSED project, 2008, Ircam, Verona, Berlin
Rolling
Clapping 1
Clapping 2Examples:
Functional ArtefactsLinking design to evaluation
including basic design methods
Sonic Interaction Design Process for
Functional Artefacts
Interaction MaterialsAnalyse,Describe
Interaction Gestalt
Experience in Context
Interaction Artefact
Shape, Combine
Evaluate
Functional Artefacts CLOSED project, 2008, Ircam, Verona, Berlin
Rolling
Clapping 1
Clapping 2Examples:
Functional ArtefactsLinking design to evaluation
including basic design methods
Sonic Interaction Design Process for
Functional Artefacts
Interaction MaterialsAnalyse,Describe
Interaction Gestalt
Experience in Context
Interaction Artefact
Shape, Combine
Evaluate
Functional Artefacts CLOSED project, 2008, Ircam, Verona, Berlin
Rolling
Clapping 1
Clapping 2Examples:
Explorations and Methods
• Listening test approaches
• Textual descriptor analysis
• Sonic Mapping
• Causal uncertainty
• 2-D perceptual scaling
• Repertory grid technique
Background Sounds Foreground Sounds
Contextual
Sounds
Sig
nals
Sig
nals
Sig
nals
Emotions
Actions
Actions
Actions
Emotions
Emotions
Visible HiddenWritten descriptions of sounds and their type
Music
Speech
Everyday
Abstract /Unknown
*+0
?
*
*+ ?
00
Sonic MappingGraham Coleman, PhD ThesisScotland, 2008
Listening
Sorting
Classifying
Background Sounds Foreground Sounds
Contextual
Sounds
Sig
nals
Sig
nals
Sig
nals
Emotions
Actions
Actions
Actions
Emotions
Emotions
Visible HiddenWritten descriptions of sounds and their type
Music
Speech
Everyday
Abstract /Unknown
*+0
?
*
*+ ?
00
Sonic MappingGraham Coleman, PhD ThesisScotland, 2008
Listening
Sorting
Classifying
1 2 3 4 5 6 7 8 9
1 2 3 4 5 6 7 8 91
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9alarmed excitedastonished
afraid delighted
angry happy
annoyed
frustrated
pleased
content
serenemiserable
depressed
bored relaxed
tired sleepy
High Arousal
Low Arousal
Dis
ple
asu
re Pleasu
re
calm
none
close/sitstroke
petting
Zizi - The Affectionate
Couch
Stephen Barrass, Linda Davy & Kerry RichensAustralia, 2003
1 2 3 4 5 6 7 8 9
1 2 3 4 5 6 7 8 91
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9alarmed excitedastonished
afraid delighted
angry happy
annoyed
frustrated
pleased
content
serenemiserable
depressed
bored relaxed
tired sleepy
High Arousal
Low Arousal
Dis
ple
asu
re Pleasu
re
calm
none
close/sitstroke
petting
Zizi - The Affectionate
Couch
Stephen Barrass, Linda Davy & Kerry RichensAustralia, 2003
Repertory Grid George Kelly,USA, 1955
Listening
Sorting
Classifying
Source / Object / Type of
Interaction
Dendogram from Repertory Grid
George Kelly,USA, 1955
Clusters of `related’ or `similar’ sounds
Advice and tips for achieving better sound designs
Selecting and classify suitable sounds
Verify your mappings
Rapid prototyping with tools such as PD, SuperCollider, Chuck, etc.
Contexts / Users / Environment