Content uploaded by Klaus Genuit
Author content
All content in this area was uploaded by Klaus Genuit on Mar 24, 2019
Content may be subject to copyright.
Dearborn, Michigan
NOISE-CON 2008
2008 July 28-30
Vehicle Interior Noise – A Combination of Sound, Vibration
and Interactivity
Klaus Genuit
HEAD acoustics GmbH
Ebertstr. 30a
52134 Herzogenrath
Germany
ABSTRACT
Human beings receive the surrounding world via different senses. Interaction phenomena are
expected in the context of sound and vibration. In fact, the evaluation of acoustical comfort in a
vehicle cannot be achieved with an isolated consideration of only the airborne noise. A passenger
must be regarded as part of a vibro-acoustic system in which the coupling with the vehicle occurs
via the contact points of the steering wheel, seat, floor panel and pedals. Within this context, and
in times of advanced acoustical engineering resulting in quiet vehicles with low interior sound
pressure levels, vibrations can become more important. The impact of combined sound and
vibration stimuli on subjective evaluations has not yet been fully clarified. In the following,
explorative studies about different test situations and their influence on evaluation results are
presented. Test subjects were asked to evaluate sound, or vibration, or both, with respect to their
quality in different test situations. Different test settings were considered: real vehicle drive,
noise and vibration reproduction in a driving simulator, and noise playback in a listening room.
The tests provided information about the interaction of sound and vibration, and about the
importance of other aspects such as context and interactivity.
1. INTRODUCTION
To guarantee a high quality of vehicle interior noise has become a very vital task for acoustic
engineers in the automotive industry for more than 30 years. However, the goals have changed
over the years. At first, acoustic engineers in the automotive industry were confronted with the
task to make the interior noise of a vehicle tolerable and to reduce the SPL of the interior noise
as much as possible. By reducing the SPL in the passenger cabin over the years, the engine
sound more and more did not mask other sounds. As a consequence, further sound sources could
be heard. Since the vehicles have continuously become quieter, the customer’s sensitiveness for
acoustical comfort and pleasant sounds increased. On the one hand, certain noise sources become
more perceptible due to a reduced overall sound pressure level. The human hearing adapts to an
average level and becomes more sensitive for any changes in the time and frequency domain.
(fig.1)
Figure 1: The adaptivity of human hearing – Permanent new challenges for acoustic engineers
Consequently, the engineers had to broaden their work scope and had to deal with more intricate
problems. They had to eliminate specific sound contributions, which were completely masked in
the past. They had to focus on several sources with different properties and characteristics.
The automotive industry realized that an explicit sound design is necessary in the field of vehicle
development in order to improve the attractiveness of the product. All in all, the sound and
vibration of a vehicle was no longer considered as only disturbing, which has to be reduced as
much as possible, but instead the drivers as well as the vehicle manufacturers noticed that an
appropriate sound is useful for driving and thereby produce driving pleasure and enhance the
attractiveness and the quality impression of the product.
2. THE INTERACTION OF SENSES
Because of permanent interaction effects between different parallel activated senses (tactile,
auditory, olfactory, visual, gustatory), single sense perception processes are combined to a
complete perception. Often, it is difficult to separate one activated sense from another without
reflecting the whole combined perception. This phenomenon plays an important role in the
context of vehicle interior noise where visual, inertial, vibrational inputs are permanently given –
whether subliminal or consciously perceivable. “We unconsciously utilize all sensory
information to evaluate sounds.”[1] Thus, in the context of noise and vibration interaction
phenomena must be expected.
Therefore, multi-modal effects in this context have to be explored to comprehend sound
perception in the presence of different sensory stimuli and to optimize vehicle sound as best as
possible reflecting further influences on the sound perception as well. Several studies have
already pointed out the necessity of consideration of both, noise and vibration, in the context of
sound quality and acoustical comfort. [2]
A passenger must be regarded as a part of a vibro-acoustic system coupled via the contact points
steering wheel, seat, floor panel and pedals with the vehicle, a coupled person-machine system.
Particularly, in times of advanced acoustical engineering resulting in quiet vehicles with low
interior sound pressure level, vibrations become more important. The noise assessments are
subconsciously influenced by the presence of other sensory stimuli, like vibration or visual
information.
Figure 2: Main contact points of vehicle drivers
A balanced configuration of the vibrational level in the vehicle has to be determined considering
comfort as well information content aspects. Therefore, further research in the field of human
perception of combined stimuli is necessary - not only for product quality reasons, but rather in
the context of occupational health and safety as well as environmental impact assessments the
presented object of investigation demands attention. It is known from studies researching the
reactions to whole-body and hand-arm transmitted vibrations that certain vibration magnitudes
and frequency components increase health risks of humans. The organs and extremities of
humans have certain resonances which could be excited e.g. via contact points within a vehicle.
However, “human responses to vibrations are varied and differ greatly over time and from one
person to the other. Therefore a vibration limit is meaningless without the specification of the
relevant criterion […]. Therefore it is difficult or impossible to summarize all effects, to define a
standard with limits and standard values for all conditions and for the whole frequency and level
range.” [3]
3. THE STUDY OF INTERMODAL ASPECTS
The results of diverse studies showed both interaction effects, vibrations have both positive and
negative effects on sound perception. Partly, it was observed that vibrations can lower the
annoyance caused by noise. [4] This effect is very interesting for automotive sound and vibration
design.
However, some studies confirmed no evidence of interaction between sound and vibration. For
example, Amman et al. assumes that there is no significant interaction of sound and vibration
and concludes “setting sound and vibration targets for vehicle programs that are independent of
one-another seems to be a reasonable approach […]. This simplifies the target setting process
significantly since sound targets do not have to be a function of vibration levels and vice-
versa.”[5] Amman at el. examines the assessment of transient NVH events that occur when
vehicle tires impact with a discontinuity in the road (impact harshness) as well as examines the
perceptual relationship of steady-state noise and vibrations experienced when driving over coarse
road.
To what extent the conclusions of Amman et al. can be transferred to all NVH-issues is very
doubtful. All in all, several publications report on interaction effects in the field of sound and
vibration. Hashimoto stated that the addition of seat and floor vibrations to vehicle interior noise
increase the perception of powerfulness, unpleasantness and booming. [6] This issue is also
intensively studied in other contexts, such as helicopters and aircrafts. [7]
Table 1 displays a selection of known concepts and theories of sound and vibration interaction
phenomena. The list shows only few widespread concepts, which are partly overlapping; but the
list is not exhaustive.
Up to now, no general concept is broadly acknowledged. Presumably, the occurrence of the
noted effects varies with the context, with the “ratio” of sound and vibration or even varies from
individual to individual. The current knowledge is insufficient to provide a complete explanation
about the interactions between these sensory dimensions, in particular when evaluating
acoustical comfort criteria.
Table 1: Concepts and theories of noise and vibration effects (partly overlapping)
Concept/Theory Short Description
Additive Interaction Vibration increase noise evaluations
Subtractive Interaction Vehicle decrease noise evaluations
Cognitive Capacity Theory Noise and vibration act as cognitive load;
in case of high cognitive load it is
difficult to separately assess the different
senses
Mismatch Theory Person mainly focuses on a single
stressor (the difference between stressors
becomes more salient)
Contrast Theory Stimuli divert attention (or mask)
effect of other stimuli
Dominating Theory One stimulus attracts more attention
than the other stimulus and
dominates the evaluation
Interaction Effects
Masking Theory e.g. high noise raise vibration thresholds
and vice versa
No Interaction Effects Assessment of combined
stimulus (effect) is equal to the sum
of individual stimuli (effects)
4. USE OF DRIVING SIMULATORS
By means of driving simulators a combined playback of sound and vibrations in real passenger
cabins is possible. Table 2 displays the importance of driving simulators with respect to
acoustical comfort and sound quality investigations in the field of automobiles.
The SoundCar is a listening environment that consists of a real vehicle cabin with authentic
control instruments and equipped with a simulation system of acoustical and vibrational
feedback. (fig. 3) The sound simulation system (H3S) provides an interactive simulation of the
current driving situation, which depends on the driver’s actions. The sound and vibration adapt to
changes in speed and engine revolutions per minute (rpm) caused by the vehicle controls. The
simulation hardware is mounted into the engine compartment. An additional subwoofer, located
in the trunk, generates low-frequency airborne sound below 150 Hz. The vibrations in the low-
frequency range are particularly important with respect to the subjective assessment of acoustical
comfort [8]. The excitation of the seat results from an electrodynamics shaker in z-direction. The
excitation of the steering takes place in the direction of rotation by means of direct control of the
steering shaft.
Table 2: Driving simulator – Some advantages and disadvantages for NVH-studies
Aspects Advantages/Disadvantages
Seat in
laboratory
(artificial
surrounding)
Exact control of noise and vibration stimuli
Conflict: artificiality of situation
Test surrounding
Realistic test
surrounding
(simulator):
Stationary
version
Mobile
version
Test person perceives stimuli in real vehicle
sitting on a original seat, seeing typical controls
and instruments, orientating in usual automotive
spatial geometry
Control of noise and vibration stimuli
Conflict: artificiality of situation, limited driving
feeling
Most realistic test surrounding
Conflict: limited comparability of the stimuli
and respective assessments
Contact points Test person experiences vibrations via seat
and/or floor panel and steering wheel
Conflict: Excitations are often realized only in
one or two DOFs over a limited frequency range
Interactivity Simulation systems reproduce noise and
vibration in dependence on test person’s actions
Conflict: limited comparability of the stimuli
and respective assessments
Room acoustics Human hearing automatically adapts to the
current location with its acoustic properties, in
case of a mismatch between noise and acoustic
room properties test persons are often confused
and irritated
The reproduction of sound at the ear is based on signals recording during bench tests of engine
noise and coast down measurements of wind and tire noise. The application of binaural
technology guarantees the realistic reproduction of the complete spatial hearing sensation. The
advantage of using headphones is the localization accuracy of the reproduction. The sound
simulation system can be used in stationary driving simulators as well as in mobile driving
simulator versions (fig. 4).
In the mobile simulator vibrations cannot be controlled and depend on the actual used vehicle, in
which the sound simulation system is installed. The advantage of a mobile sound simulation is
that a perfect driving dynamic model and an authentic vibrational setting are given through the
usage of the real vehicle. In contrast, a stationary sound simulation allows the control for
example of the vibrational stimulus, but the authenticity of the driving situation is reduced. With
respect to the object of investigation the necessary test environment has to be selected.
Figure 3: Stationary driving simulator (SoundCar); left: interior view, right: exterior view
Figure 4: Mobile driving simulator equipped with 3D Sound Simulation System (H3S), left: Test person, right:
Experimental leader
5. INVESTIGATIONS OF NOISE AND VIBRATION EFFECTS
An explorative study was carried out investigating the influence of vibrations on evaluations of
acoustical comfort criteria using a stationary driving simulator (SoundCar). Five different
vehicles (station wagons) in various operating modes were recorded. The binaural signals (the
receiver position was the driver position) as well as the acceleration of the seat in z-direction and
the acceleration of the steering wheel in the direction of rotation were available for the listening
tests. The test persons had to rank the stimuli of a specific operating mode between 1 (worst) and
5 (best), equal ratings were also accepted. Three runs were carried out, in the first run the interior
noise was offered together with the corresponding vibration excitations, in a second run only the
airborne noise contributions of the five vehicles were presented. The stimuli were presented
while the test persons were sitting in the stationary driving simulator. In a third run, the test
persons ranked the different noise stimuli in the laboratory. The runs took place at different days
to exclude the influence of memory effects. In the following some results of this study are
depicted.
50km/h in 3rd Gear
0,0
1,0
2,0
3,0
4,0
5,0
6,0
A B C D E
Vehicle
Ran king
N&V - SoundCar N - SoundCar N - Lab
50km/h in 3rd Gear
0,0
0,5
1,0
1,5
2,0
2,5
3,0
A B C D E
Vehicle
Differe nce
Dif N&V (SoundCar) - N (SoundCar ) Dif N&V (SoundCar ) - N (Lab)
Dif N (SoundCar) - N (Lab)
Figure 5: Rankings of five vehicles in the operating mode 50km/h in 3
rd
gear; evaluation criterion: appropriate noise
to slow driving (left); Differences in rankings between the settings (right)
3500 rp m in 2nd gear
0,0
1,0
2,0
3,0
4,0
5,0
6,0
A B C D E
Vehicle
Rank ing
N&V - SoundCar N - SoundCar N - Lab
3500 rp m in 2nd gear
0,0
0,5
1,0
1,5
2,0
2,5
3,0
A B C D E
Vehicle
Differe nce
Dif N&V (SoundCar) - N (SoundCar ) Dif N&V (SoundCar ) - N (Lab)
Dif N (SoundCar) - N (Lab)
Figure 6: Rankings of five vehicles in the operating mode 3500rpm in 2
nd
gear; evaluation criterion: acoustical
comfort (left); Differences in rankings between the settings (right)
Starting the eng ine
0,0
1,0
2,0
3,0
4,0
5,0
6,0
A B C D E
Vehicle
Rank ing
N&V - SoundCar N - SoundCar N - Lab
Starting the eng ine
0,0
0,5
1,0
1,5
2,0
2,5
3,0
ABCDE
Vehicle
Differe nce
Dif N&V (SoundCar) - N (SoundCar ) Dif N&V (SoundCar ) - N (Lab)
Dif N (SoundCar) - N (Lab)
Figure 7: Rankings of five vehicles in the operating mode starting the engine; evaluation criterion: perceived
acoustical quality (left), Differences in rankings between the settings (right)
The figures 5, 6 and 7 on the left side show the averaged ranks of each vehicle in different
operating modes assessed in different test surroundings and conditions. The standard deviation is
also depicted in the diagrams. Furthermore, the figures 5, 6 and 7 on the right side display the
differences of the averaged ranks between the different test conditions. The presented noise was
identical; the vibrations were only presented in scenario 1.
It was observed that the evaluations differ depending on the vibration presentation. Even
between the test surroundings SoundCar and laboratory, where the stimuli – only airborne noise
– were identical, small evaluation differences could be observed. It has to be remarked that the
evaluation criteria partly were not exclusively noise-related. For example, the evaluation
criterion “appropriate to slow driving” allow for the intentional consideration and integration of
the vibration magnitude into the assessment. However, two conclusions can be drawn from this
case study. First, the evaluation differences between noise playback within SoundCar and
laboratory (red bars) show the influence of the test surrounding and context on the evaluations.
The judgments, although the noise stimuli are identical, are not completely equal. Second, the
presence and magnitude of vibrations effect the evaluation of the perceived noise.
Of course, the ranking method does not allow definite conclusions. The different rankings do not
offer any clues with respect to the exact cause and quantitative extent of the numerical
differences. Nevertheless, the study shows the importance of vibrations to noise assessments and
the importance of the context (realistic vs. artificial surrounding).
Another experiment was conducted to study effects caused by the selection of the test
environment. Evaluations of sound and vibration stimuli were given during a real test drive on
the road and during driving a simulator.
In the driving simulator no moving scenery information or other visual input, except the test
environment itself, was provided. Different operating modes of six vehicles were assessed with
respect to overall quality, quality of engine noise, tire noise, wind noise, seat vibration and
steering wheel vibration using a scale corresponding to school grades, ranging from
excellent/very good (1) to inadequate/fail (6). Eleven test subjects took part in each test scenario,
the test persons were experienced very well with sound and vibration evaluation tasks. The
judged driving conditions were 4
th
gear at a constant speed 100km/h and 130km/h, 2
nd
gear run-
up with defined load condition and an engine startup.
Exemplarily, some results of the evaluations given in a real car compared with evaluations made
in the stationary driving simulator are depicted in the following diagrams. (fig. 8 and 9)
First of all, the tests have shown that good correlation exists between NVH-evaluations made
during test drives in a real vehicle and evaluations made in a driving simulator of reproduced
sounds and vibrations. Small evaluation differences could be observed concerning the
evaluations of the engine noise quality in the real car drive and in the simulator drive as well as
considering the different evaluations of the steering wheel vibrations. The given evaluations
varied in ½, 1 up to 2 grades.
The ratings in the real car differ partially from the driving simulator assessments. Presumably,
the driving task in reality has attracted more attention and the inadequateness of the stimuli was
perceived as less apparent. However, this effect cannot be generalized because of the small test
group. It can only be concluded that evaluations of noise and vibration are context-dependent
and are influenced by the respective test environment.
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
Rating
TP 1 TP 2 TP 3 TP 4 TP 5 TP 6 TP 7 TP 8 TP 9 TP 10 TP 11 Mea n
Test Person
Engine Noise - Comparison of Real Car to Driving Simulator Evaluations - 100km/h, 4th gear
Real Car SoundCar (Driving Simulator)
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
Rating
TP 1 TP 2 TP 3 TP 4 TP 5 TP 6 TP 7 TP 8 TP 9 TP 10 TP 11 Me an
Test Person
Steering Wheel Vibr ations - Comparis on of Real Car to Driving Simulator Evaluations - 100km/h, 4th Gear
Real Car SoundCar (Driving Simulator)
Figure 8: Comparison of ratings: one vehicle in operating mode constant speed 100km/h (4
th
gear) in the real car
and played back in a driving simulator (N&V); top: engine noise quality; below: steering wheel vibration quality
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
Rating
TP 1 TP 2 TP 3 TP 4 TP 5 TP 6 TP 7 TP 8 TP 9 TP 10 TP 11 Mean
Test Person
Engine Noise - Compa rison of Real Car to Driving Simulator Evalua tions - 130km/h, 5th Gear
Real Car SoundCar (Driving Simulator)
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
Rating
TP 1 Tp 2 TP 3 TP 4 TP 5 TP 6 TP 7 TP 8 TP 9 TP 10 TP 11 Mea n
Test Pe rson
Stee ring Wheel V ibrations - Comparison of Real Car and Driv ing Simulator Evaluations - 130 km /h, 5th Gear
Real Car Sound Car (Driving Simulato r)
Figure 9: Comparison of ratings: one vehicle in operating mode constant speed 130km/h (5
th
gear) in the real car
and played back in a driving simulator (N&V); top: engine noise quality; below: steering wheel vibration quality
6. INTERACTIVITY AND EXPLORATIVE METHODS
Listening tests are usually carried out to collect data about subjective evaluations of predefined
stimuli. For reproducibility and analysis reasons the tests are often completely standardized and
conducted in a “controlled” test surrounding. Such test conditions should allow the provable
statistical correlation between subjective ratings and objective parameters. However, the abstract
test situation, the stimuli representation taken out of the typical context, and the response
limitation (pressing complex sensations into given scales) leads to biased results, which often
cannot be confirmed in reality. Aspects, such as context, ambiance, interactivity, occurrence of
several sensory inputs (combined stimuli), moderate the perception and evaluation of (noise)
stimuli.
The evaluation of vehicle noise is also dependent from several aspects. The evaluation of vehicle
interior noise quality is based on the resulting vibro-acoustic exposure. Thus, the application of
new types of listening tests regarding reality-relevant aspects, e.g. the method EVE (Explorative
Vehicle Evaluation), seems imperative. [9]
The EVE method considers the context, where evaluations of vehicle sounds are usually carried
out, the interactivity, which occurs between the driver and the vehicle in common driving
situations, the spontaneity of evaluations in real drive. Within the method the process of
evaluating vehicle sounds is relocated to a driving vehicle (mobile driving simulator). An
advantage is that the test person creates their own stimuli, acts on their own initiative to evaluate
the heard sounds and can explain his/her feelings and judgments in an open and free way not
confined to give forced answers to predefined quantities and scales. Therefore, the evaluation of
sound events interactively takes place and is modified by its significance in and for the ambient
space.
The analysis of the comments and the stimuli which the comments refer to is done with a
qualitative analysis technique (e.g. Grounded Theory) as well as with triangulation methods.
Apart from the verbal evaluations, the commented binaural signal, speed, gear, rpm information
as well as seat vibrations are recorded. The analyses of previous EVE-studies showed that few
noise evaluations given by naive test persons were subconsciously induced by vibrations.
The technique gives insights into the perception and evaluation of vehicle sounds. The purpose
behind the analysis is to group similar events, sensations, reactions under a common heading or
classification as well as to discover intersubjective evaluation patterns. Based on the results
information concerning the character and features of a preferred vehicle sound is derived. On the
basis of the tests a vehicle sound is developed considering the different remarks, customer’s
preferences and acoustical analyses results. The target sound must be also matched with the
vibrational conditions of the investigated vehicle. Moreover, disturbing vibrational components
have to be identified to avoid negative acoustical comfort evaluations mainly caused by the
vehicle vibrations.
7. CONCLUSIONS
Sound and vibration must not be only considered as disturbing elements regarding the NVH-
comfort; they could be used to develop an acoustical and vibrational environment which
influences positively the customer’s contentment. The creation of sounds, which result in an
impression of high product quality, helps manufacturers to stand out against their competitors
and to meet increased requirements. Sound quality studies considering both sound and vibration
require multi-dimensional approaches. All in all, further research is needed in the field of human
perception of combined stimuli. The existing literature and the presented studies show the
complexity of interaction effects caused by sound and vibration in vehicles. Some studies even
show no (or little) evidence for interaction effects; this means that the assessment of the
combined stimuli (effect) is almost equal to the sum of individual stimuli assessments. This
contradiction can only be solved with the help of further research.
The use of a driving simulator, whether stationary or mobile, provides an opportunity to control
the stimuli while at the same time creating an interactive experience of the whole vehicle,
including its sound, vibration, haptic and visual information.[10] Simulators allow for a focused
investigation of multi-sensory effects under authentic test conditions, permitting the investigation
of interaction effects, and leading to the development of methods that take into account comfort-
oriented NVH-engineering, including cross-modal perception phenomena.
The topic of combined stimuli is not only of importance to the field of product quality
optimization but also within the context of occupational health and safety as well as
environmental impact research.
REFERENCES
1
Abe, K., Ozawa, K., Suzuki, Y., Sone, T., “Comparison of the effects of verbal versus visual information about
sound sources on the perception of environmental noise”, Acta Acustica, Vol. 92, pp. 51-60 (2006).
2
Genuit, K., “The interaction of noise and vibration inside vehicles”, 8
th
International Congress on Sound and
Vibration, Hong Kong, China (2001).
3
Bellmann, M., “Perception of whole-body vibrations: from basic experiments to effects of seat and steering wheel
vibrations on the passenger’s comfort inside vehicles, dissertation, University of Oldenburg, Germany (2005).
4
Genell, A., Västfjäll, D., “Vibrations can have both negative and positive effects on the perception of sound”, Int. J.
Vehicle Noise and Vibration, Vol. 3, No. 2 (2007).
5
Amman, S., Mouch, T., Meier, R., “Sound and vibration perceptual contributions during vehicle transient and
steady-state road inputs”, Int. J. Vehicle Noise and Vibration, Vol. 3, No. 2 (2007).
6
Hashimoto, T., “Tradeoff level of the visual scenery and seat/floor vibrations to the perception of sound quality of
car interior noise”, Inter-Noise 2004, Proceedings, Prague, Czech Republic (2004).
7
Mellert, V., Baumann, I., Freese, N., Weber, R., “Investigation of noise and vibration impact on aircraft crew,
studied in an aircraft simulator”, Inter-Noise 2004, Proceedings, Prague, Czech Republic (2004).
8
Giacomin, J, Woo, Y.J., “Beyond comfort: information content and perception enhancement”, Engineering
Integrity, Vol. 16, July, pp. 8-16 (2004).
9
Schulte-Fortkamp, B., Genuit, K., Fiebig, A., “New approach for the development of vehicle target sounds”, Inter-
Noise 2006, Proceedings, Honolulu, HI, USA (2006).
10
Genuit, K., “Interactive simulation tools for the investigation of multi-sensory effects, Inter-Noise 2006,
Proceedings, Honolulu, HI, USA, (2006).
