The Effect of Weight on the Perception of Vibrotactile Intensity with Handheld Devices.

Full text

Page 1

The Effect of Weight on the Perception of
Vibrotactile Intensity with Handheld Devices
Hsin-Yun YaoVincent Hayward
Haptics Laboratory, Ctr. For Int. Machines,
McGill Univ., Montr´ eal, Qc, H3A 2A7 Canada
{hyyao,hayward}@cim.mcgill.ca
Manuel CruzDanny Grant
Immersion Canada Inc.,
Montr´ eal, Qu´ ebec, H2W 2R2 Canada
{mcruz,dgrant}@immersion.com
Abstract
The objective of this study was to determine whether the
weight of a vibrating handheld object influenced the per-
ceived intensity of its vibrations. Experiments were con-
ducted to determine the subjective equivalence of vibrotac-
tile intensity for objects that had the same size but had dif-
ferent weights. The results suggest that for the same surface
acceleration and hence the same movement, the heavier is
the device, the stronger is the perceived intensity.
1Introduction
Vibrotactile signals are nowadays important for the de-
sign of a variety of handheld devices. To design them, it
is important to know whether the perceived vibrotactile in-
tensity depends on other object attributes, chiefly among
them is weight. Despite a long history of vibrotactile stud-
ies we could find little data that was related to this ques-
tion [1,4–6,8].
Most portable phones provide a vibrotactile function to
signal a call. Design engineers must carefully select the
actuator to fit a tight power budget. We observed however
that the way a device felt in our hand seemed to depend
on its weight, and this despite compensating for the physics
of vibration. This led us to hypothesize the existence of a
weight-vibration perceptual interaction.
We carried out an experiment to determine the subjec-
tive equivalence of the vibration intensity for objects having
different weights. We found that for the same acceleration,
doubling the weight of an object resulted a perceptual sen-
sitivity enhancement of about 2.4 dB.
2Method
Wemanufacturedboxeswithweightsandsizessimilarto
that of common portable phones. Each was equipped with
a high-bandwidth actuator and an accelerometer attached to
its surface. The method of adjustment was used whereby
participants adjusted the vibrating intensity of a given box
to match that of a reference box.
Apparatus.
their controlling circuitry, were made. The boxes weighted
50, 110, 200 g. A fourth 110 g box served as the reference.
Each contained a custom-made, recoil-type Lorentz actua-
tor made of a magnet suspended inside a pair of coils in
an open magnetic circuit arrangement. It could accelerate
a 110 g box up to 30 m/s2from 20 to 500 Hz with min-
imal distortion. The intensity was adjustable by turning a
rotary control similar to that of audio equipment. Each had
a calibrated accelerometer (MMA7260Q, Freescale).
Four boxes (80 × 40 × 17 mm), along with
Stimulus.
amplitudes from 3 to 14 m/s2. The signals were pulsed in a
0.5-second-on, 0.5-second-off duty cycle to minimize adap-
tation. The frequency, magnitude and the pulse-train signal
approximated that of an actual a portable phone.
We used sinusoidal vibrations of 150 Hz with
Protocol.
males) were recruited. During each session, the participants
were seated in front of the four boxes which rested each
on a block of soft gel. They were told that there was one
reference box and 3 adjustable boxes. For each trial, the
reference box and one of the tunable boxes would vibrate.
They had to lift, feel the boxes, and then adjust the inten-
sity of the tunable box until its vibration felt the same as the
reference box. They wore sound-blocking earphones and
used their dominant hand only. When a participant was sat-
isfied, she put both boxes back on their block of gel. The
acceleration of each box was recorded and logged by com-
puter before the next trial. Each participant was presented
20 pairs of stimuli in total, and they were asked to take a
one-minute break after the 12th pair. All completed the ex-
periment within 30 minutes.
Eight healthy university students (4 males, 4 fe-
Proceedings of World Haptics 2007 (The Second Joint Eurohaptics Conference And Symposium
On Haptic Interfaces For Virtual Environment And Teleoperator Systems), pp. 551-552.

Page 2

3 Results
The results (acceleration r.m.s. converted to dB re 10−6
m/s2), see Fig. 1, showed that the weight of the device in-
fluenced the perception of vibration magnitude. The data
for each condition was fitted with straight lines. The con-
trol condition was when participants had to match intensity
of boxes of same weight. On average, they behaved almost
like the ideal performer (thick line) for this condition. We
calculated the normalized acceleration by taking the ratio
of the matched acceleration over the reference acceleration.
The ANOVA test performed for each of the three pairs of
data (50 g vs 110 g, 110 g vs 200 g, 50 g vs 200 g) showed
significant difference for all the pairs (p <0.001). On the
boxplot, the notch indicates a robust estimate of the uncer-
tainty about the medians for box-to-box comparison. Since
the notches do not overlap, the plot indicates the medians
of each pair differ at the 5 % significance level. The re-
sults show that a heavier box requires less acceleration to
produce the same perceptual effect than a lighter one.
8
6
4
2
0
-2
-4
-6
50 g
110 g
220 g
140
110
120
130
121127 130133
Reference signal in dB
Matched signal in dB
50 g
110 g
220 g
Matched acceleration
Reference acceleration
(dB, re = 10 m/s )
-62
124118
Figure 1. Matched values vs reference values.
4Discussion
Our results support the hypothesis that the weight of
an object affects the perception of vibration magnitude.
The heavier the box, the smaller is the required accelera-
tion to produce the same perceptual intensity. One possi-
ble explanation is that when we hold a heavy box in our
hands, we need to use a stronger grip, therefore the con-
tact area between the skin and the device is larger than with
a lighter box [7], which, in turn, stimulates a larger number
ofmechanoreceptors. Asecondexplanationisthataheavier
object can cause more tissues to vibrate for the same accel-
eration. A third explanation would appeal to psychophys-
ical mechanisms [3]. Whatever the biomechanical, neuro-
physiological, or psychological factors may be, our nervous
system translates this vibration pattern into the perception
of stronger intensity.
The weights were specifically selected to have almost
the 1:2:4 ratios. As seen in Fig. 1, the relative perceptual
differences between the successive weights are almost the
same: 2.4 dB between 50 g and 110 g, and 2.3 dB between
110 g and 200 g. This difference was relatively constant
for all signal amplitudes within our range of testing. This
means that, on average, for the range of weight between 50
and 200 g, to obtain the subjective equivalence for a device
twice as heavy, one need to reduce by about 2.4 dB to the
response of the original box.
5Conclusion
We have found that the design of the vibrotactile signal
given by a handheld device should take its weight into con-
sideration. This result is potentially useful for the portable
device industry as it provides a basic guideline for vibrotac-
tile tactile transducers.
It tempting to relate our findings to the well-known size-
weight illusion [2]. Our present protocol makes an implicit
causal connection between weight and the perception of vi-
brations as it is motivated by our application. It would be
interesting to investigate whether a reverse interaction oc-
curs.
References
[1] A. J. Champagne. Correlation of electric power steering vi-
bration to subjective ratings. SAE Technical Paper Series,
2000-01-0176:1–3, March 6–9 2000.
[2] A. Charpentier. Analyse exp´ erimentale de quelques ´ el´ ements
de la sensation de poids. Archives de Physiologie Normale et
Pathologique, 3:122–135, 1891.
[3] J. C. Craig and C. E. Sherrick. The role of skin coupling in the
determination of vibrotactile spatial summation. Perception
and Psychophysics, 6:97–101, 1969.
[4] D. B. Fleming and M. J. Griffin. A study of the subjective
equivalence of noise and whole-body vibration. Journal of
Sound and Vibration, 42(4):453–461, 1975.
[5] G. A. Gescheider, J. S. J. Bolanowski, R. T. Verrillo, D. J.
Arpajian, and T. F. Ryan. Vibrotactile intensity discrimination
measured by three methods. Journal of the Acoustical Society
of America, 87(1):330–338, 1990.
[6] S. Maeda.Necessary research for standardization of sub-
jective scaling of whole-body vibration. Industrial Health,
43(3):390–401, 2005.
[7] R. T. Verrillo. Effect of contactor area on the vibrotactile
threshold.Journal of the Acoustical Society of America,
35(12):1962–1966, 1963.
[8] P. G. Wiles, S. M. Pearce, P. G. Rice, and J. M. Mitchell.
Vibration perception threshold: influence of age, height, sex,
and smoking, and calculation of accurate centile values. Dia-
betic Medicine, 8(2):157–161, 1991.
Proceedings of World Haptics 2007 (The Second Joint Eurohaptics Conference And Symposium
On Haptic Interfaces For Virtual Environment And Teleoperator Systems), pp. 551-552.