- Email: [email protected]
Velocity judgments of continuously moving and stroboscopically presented stimuli using magnitude and category scaling
Acta Psychologica 48 (I 98 I) 89- 94 North-Holland Publishing Company
89
VELOCITY JUDGMENTS OF CONTINUOUSLY MOVING AND STROBOSCOPICALLY PRESENTED STIMULI USING MAGNITUDE AND CATEGORY SCALING F.P. VAN EYL * Hope College, MI, USA
An experiment was conducted to determine whether perceived velocities of continuously moving targets and stimuli presented stroboscopically (in successively different positions) are processed by two different motion-sensitive mechanisms: one substitutive (metathetic), the other additive (prothetic). Male subjects made category scale and magnitude scale judgments of targets presented continously and stroboscopically at five different velocities. When the magnitude scale data were plotted against the category scale data, linear relationships emerged, suggesting that in both conditions velocity perception is due to a substitutive process. The results are discussed in view of previous research and theoretical implications.
There is considerable evidence from studies of single-cell response and psychophysical research on direction-specific adaptation to suggest that velocity detectors exist (e.g., Richards 197 1; and Sekuler et al. 1978). It has also become apparent that individual detector units have degrees of velocity specificity. Less clear is whether the perception of physically moving objects and apparent movement, specifically /3 movement, are processed in similar fashion by the same mechanism. The question stems from an observed time differential between the perception of actual movement and stroboscopic movement. Lit (1960), for instance, using the Pulfrich pendulum found a considerably shorter lag time than Roufs (1974) who used a double-flash method. Although this difference (13 us 30 msec) was initially attributed to a luminance factor, Wagenaar et ai. (1980a, b) demonstrated that a stroboscopically presented target is consistently perceived as lagging behind a continuously moving target and the lag time is not so much a function of luminance as it is of velocity and flicker frequency. They found that lag time increases as velocity increases and flicker frequency decreases. Because the data did not fit the persistence-of-an-after-image model either, the existence of two mechanisms seemed more likely. * Author’s
address:
Dept. of Psychology,
0 00 l-69 18/8 1/OOOO-0000/$02.50
Hope College, Holland,
MI 49423, USA.
0 198 1 North-Holland
Following up on Gregory’s (1964) proposal that velocity is processed directly when the stimulus is continuous and indirectly or inferred when the stimulus is shown in successive positions, Wagenaar et al. (1980a) suggested two different coding systems. “The motion of continuously visible stimuli is processed directly, by means of a specific motion-sensitive mechanism which we will call ‘velocity detection system’. The movement of flickering stimuli could be mediated by the system responsible for phenomena of apparent movement, specifically p movement. We call this mechanism ‘stroboscopic system”’ (1980a: 9). If indeed two different processes exist, it is also possible that one process has a qualitative, substitutive or metathetic dimension, the other a quantitative, additive or prothetic dimension. In view of the fact that there is evidence for velocity specificity (Sekular et al. 1978), it is most likely that the velocity detector cells function in a qualitative, substitutive fashion or metathetically, analogous to the way the cells along the basilar membrane function. If so, the process of velocity judgment, when plotted logarithmicly, should manifest itself as a linear function. The change-of-place or stroboscopic stimulus presentation could be quantitative or prothetic in process which, if true, should manifest itself as a non-linear function (Stevens and Galanter 1957) when plotted logarithmicly. According to Stevens (1975) it is often possible to classify a stimulus continuum by plotting the results of the magnitude estimation scaling technique against the results obtained with category scaling. A non-linear relationsip means a prothetic continuum with an additive mechansism, while a linear relationship points to a metathetic continuum of a substitutive system. To find support for a dual-process theory of motion perception Wagenaar and Vander Schaaf (1980), used continuously moving and stroboscopically presented stimuli, subjects that used magnitude estimation and category scaling techniques, and Stevens’ (1975) contention, as stated above, that additive and substitutive mechanisms yield different functional relationships. The results were inconclusive and, if anything, showed a tendency for a non-linear relationship for both stimulus conditions. The present experiment is a second attempt using the same scaling techniques, continuous and stroboscopic stimulus presentations, but with fewer velocities, twice as many judgments, and the presentation of the standard stimulus or range stimuli before every trial.
The experiment
Subjects The Ss were twelve male volunteers
from introductory, undergraduate classes. They had no previous experience in psychology experiments.
psychology
F. P. Vun Eyl /
Veloclt~ judgments
91
Apparatus The stimulus was a projected light dot, 3 cm in diameter (visual angle 24’), moving horizontally at eye level, from left to right, a distance of 170 cm (visual angle 21”). A stationary red dot, also 3 cm in diameter, serving as a fixation point, was 16 cm above the stimulus trajectory and in the S’s objective median plane. S was seated on a chair 457 cm from the screen on which the experimental and fixation stimuli were projected. Projection of the moving and fixation dots was in darkness and accomplished with two Kodak Carousel projectors. The white stimulus moved across the screen by way of a rotating mirror. Its travel distance was limited to 170 cm by a screen with a slit. Velocity of the moving stimulus was regulated with a rheostat. The stroboscopic effect was obtained by placing a rotating disc with a slit in the path of the light beam. The speed of the disc was chosen so that the stimulus was visible for 5 msec with a 5 Hz flicker frequency. Because stroboscopic light has less brightness than the same source presented continuously, the projection of the continuous stimulus was filtered (Kodak Wratten Filter No. 96, N.D. 0.10) to make the brightness of both conditions subjectively equal. Procedure Six Ss were used for the continuous light condition and six for the stroboscopic light condition. Under both conditions the stimulus was presented at speeds of 4, 8, 12, 16 and 20 deg/sec, four times, in random order. Half the Ss were instructed to use a magnitude scale first and then, after a IO-min rest, a category scale. The other half of the Ss used the two scaling techniques in opposite order. Scaling order was randomized over Ss. For both scaling techniques every stimulus presentation was preceeded by a standard medium speed (magnitude estimation), low speed stimulus and a high speed stimulus to indicate the range (category scale). In the case of magnitude estimation the standard was 100, for category scaling the fast stimulus was called 1 and the slow one 7. As the standard was presented E announced the value “100” and requested S’s estimate of the “next one”. During the category scaling condition E announced the values “1” and “7” or “7” and “1” which was also followed by a request for a value on the “next one”. During the experiment S was also instructed repeatedly to remain looking at the red fixation point and not to track the white dot. S was alerted to the appearance of each light by E saying “now”.
Table I
Mean judgments per rating scale, velocity condition, and light stimulus presentation. Velocity
deg/sec
Continuous light Stroboscopic light
Category
Magnitude
scale
estimation
20
16
12
8
4.
20
16
I2
8
4
6.8 6.6
6.3 5.6
4.6 4. I
2.8 3.2
1.5 1.3
135 139
124 I21
106 103
71 83
52 58
I
765-
I
1
. 0
P
1
I
I
. continuous o stroboscopic
:
L0 .
32-
. 0 I
I
I
L
8
12
velocity
I”
i
16 20
deg/sec
Fig. I. Velocity judgments of continuous (solid dot) and stroboscopic obtained with a category scaling technique.
(open dot) movement
as
Results All judgments tiere averaged over the Ss per continuous movement/category scaling condition, continuous movement/magnitude estimation condition, stroboscopic presentation/category scaling condition, and stroboscopic presentation/magnitude estimation condition. The results are presented in table 1 and graphically shown in figs. I-3. Figs. 1 and 2 show the judgments by category scale and magnitude estimation,
I
20( I-
continuous 7 - o stroboscopic l
15( )-
8
1O(I-
a
Is
z
.
ct c6 ,5 -t
1
.
a
5( l-
lo
0
5
a
: :
1
1
1
continuous o stroboscopic l
0
E 1
1 1
velocity
8 m
16 20
40 mean
deg/sec
Fig. 2. Velocity judgments of continuous obtained with magnitude estimation. Fig. 3. Magnitude estimations plotted against each other.
12
100 magnitude
(solid dot) and stroboscopic
and category
ratings
I
I
IO,
L
of continuous
(open
and
150 estlmatlon
dot) movements
stroboscopic
as
movement
F.P. Vun &vl /
Velocitv judgments
93
respectively. Fig. 3 has the category ratings plotted against those obtained with magnitude estimation. It is observed that differences between the continuous light data and stroboscopically obtained data are quite small and that all graphs have an unmistakably linear character. Of course, the logarithmic transformations in figs. 1 and 2 may contribute to the apparent linearity. Nevertheless, linear regression accounts for 78 percent of the variance of the mean category judgments and 76 percent of the mean magnitude estimation judgments for the continuous condition. Linear regression also accounts for 82 percent of the variance of the mean category judgments and 84 percent of the mean magnitude estimation judgments for the stroboscopic condition. The linear correlation between the magnitude estimation and the category scale means is r=0.80 for the continuous condition and r=0.78 for the stroboscopic condition. Since the data of both continuously and stroboscopically presented stimuli have linear continua, the dimensions are metathetic which leads to the conclusion that the mechanism or mechanisms by which they were processed are substitutive in nature.
Discussion In the present experiment perceived velocity following both continuous and stroboscopically presented stimuli was linearly related to physical velocity. This strongly suggests the possibility of a substitutive process for both dimensions. Wagenaar and Vander Schaaf (1980) on the other hand, found some inconclusive indications of non-linearity, suggesting the possibility of an additive process for both conditions. Although the present experiment limited itself to just one flicker frequency (5 Hz), this cannot be the source of discrepant findings because this frequency was the same one used by Wagenaar and Vander Schaaf (1980). A more likely source is the greater reliability of the present data because twice as many judgments were taken and the standard stimulus or the range was presented before every trial. Theoretically, the results of the present experiment do not support the possibility that continuous movement and movement as a result of abrupt, discontinuous changes of location differ because one adheres to a qualitative process, the other to a quantitative one. Unless Stevens’ (1975) rationale is faulty or does not apply to the present experiment, two possibilities suggest themselves: (1) there are two separate mechanisms but both are metathetic, or (2) there is just one mechanism and any dual-process notions should be abandoned. In so far as motion is defined as displacement in space, a truism about motion is that it always has direction. Quite conceivably velocity and direction are neuro-physiologically interrelated in such a way that information processed by velocity and direction detectors combine and interact at a higher level, If direction detection is contingent on changes in location, either because the background shifts against the target or the target shifts against the background, an uninterrupted stream of successive and small location changes is processed
more efficiently than a constantly interrupted flow of information about larger (more different) location changes. Particularly if it can be assumed that velocity detectors are non-functional during very short (5 msec) stimulus presentations in successive but different locations, it leaves any higher interpretive mechanism dependent on information of frequency and extent of location changes only. Of the two, frequency analysis is more likely to be the process that functions substitutively and, as such, the primary factor of inferred velocity during stroboscopic movement. Another possibility is that the primary factor is spatio-temporal modulation sensitivity and that the spatial and temporal separation ratios of stroboscopic presentations are processed substitutively. However, regardless of the outcome, answers can be expected to be within the context of processes that are substitutive rather than additive.
References Gregory, R.L.. 1964. Human perception. British Medical Bulletin, section 4: Perception of movement, 23-25. Lit. A., 1960. The magnitude of Pulfrich stereo phenomenon as a function of target velocity. Journal of Experimental Psychology 59, l65- 175. Richards, W.. 1971. Motion detection in man and other animals. Brain Behavior and Evolution 4. 162-181. Roufs, J.A., 1974. Dynamic properties of vision: V. Perception lag and reaction time in relation to flicker and flash thresholds. Vision Research 14 (9). X53-869. Sekuler, R.W.. A. Pantle and E. Levison, 1978. ‘Physiological basis of motion perception’. In: R. Held, H. Leibowitz and H.L. Teuber (eds.), Handbook of sensory physiology. VIIII. Berlin: Springer-Verlag. pp. 67-96. Stevens, S.S., 1975. Psychophysics (edited by G. Stevens). New York: Wiley. Stevens, S.S. and E.H. Galanter. 1957. Ratio scales and category scales for a dozen perceptual continua. Journal of Experimental Psychology 54, 377-41 I. Wagenaar, W.A. and T.W. vander Schaaf, 1980. SubJective anelheidsschalen van continue en stroboscopische stimuli. Report of the Institute for Perception, RVO-TNO. Soeaterbcrg. Netherlands. Wagenaar, W.A.. T.W. vander Schaaf and G.B. Flares d’Arcais. IYHOa. An experimental study of MacKay’s illusion: evidence for two movement perception systems. Report of the Institute for Perception, RVO-TNO, Soesterberg, Netherlands. Wagenaar, W.A.. S. Luthra and G.B. Florea d’Arcais, 1980b. Effects of luminance and velocity on perception lag in moving targets. Report of the Institute for Perception, RVO-TNO, Soestcrberg, Netherlands.
89
VELOCITY JUDGMENTS OF CONTINUOUSLY MOVING AND STROBOSCOPICALLY PRESENTED STIMULI USING MAGNITUDE AND CATEGORY SCALING F.P. VAN EYL * Hope College, MI, USA
An experiment was conducted to determine whether perceived velocities of continuously moving targets and stimuli presented stroboscopically (in successively different positions) are processed by two different motion-sensitive mechanisms: one substitutive (metathetic), the other additive (prothetic). Male subjects made category scale and magnitude scale judgments of targets presented continously and stroboscopically at five different velocities. When the magnitude scale data were plotted against the category scale data, linear relationships emerged, suggesting that in both conditions velocity perception is due to a substitutive process. The results are discussed in view of previous research and theoretical implications.
There is considerable evidence from studies of single-cell response and psychophysical research on direction-specific adaptation to suggest that velocity detectors exist (e.g., Richards 197 1; and Sekuler et al. 1978). It has also become apparent that individual detector units have degrees of velocity specificity. Less clear is whether the perception of physically moving objects and apparent movement, specifically /3 movement, are processed in similar fashion by the same mechanism. The question stems from an observed time differential between the perception of actual movement and stroboscopic movement. Lit (1960), for instance, using the Pulfrich pendulum found a considerably shorter lag time than Roufs (1974) who used a double-flash method. Although this difference (13 us 30 msec) was initially attributed to a luminance factor, Wagenaar et ai. (1980a, b) demonstrated that a stroboscopically presented target is consistently perceived as lagging behind a continuously moving target and the lag time is not so much a function of luminance as it is of velocity and flicker frequency. They found that lag time increases as velocity increases and flicker frequency decreases. Because the data did not fit the persistence-of-an-after-image model either, the existence of two mechanisms seemed more likely. * Author’s
address:
Dept. of Psychology,
0 00 l-69 18/8 1/OOOO-0000/$02.50
Hope College, Holland,
MI 49423, USA.
0 198 1 North-Holland
Following up on Gregory’s (1964) proposal that velocity is processed directly when the stimulus is continuous and indirectly or inferred when the stimulus is shown in successive positions, Wagenaar et al. (1980a) suggested two different coding systems. “The motion of continuously visible stimuli is processed directly, by means of a specific motion-sensitive mechanism which we will call ‘velocity detection system’. The movement of flickering stimuli could be mediated by the system responsible for phenomena of apparent movement, specifically p movement. We call this mechanism ‘stroboscopic system”’ (1980a: 9). If indeed two different processes exist, it is also possible that one process has a qualitative, substitutive or metathetic dimension, the other a quantitative, additive or prothetic dimension. In view of the fact that there is evidence for velocity specificity (Sekular et al. 1978), it is most likely that the velocity detector cells function in a qualitative, substitutive fashion or metathetically, analogous to the way the cells along the basilar membrane function. If so, the process of velocity judgment, when plotted logarithmicly, should manifest itself as a linear function. The change-of-place or stroboscopic stimulus presentation could be quantitative or prothetic in process which, if true, should manifest itself as a non-linear function (Stevens and Galanter 1957) when plotted logarithmicly. According to Stevens (1975) it is often possible to classify a stimulus continuum by plotting the results of the magnitude estimation scaling technique against the results obtained with category scaling. A non-linear relationsip means a prothetic continuum with an additive mechansism, while a linear relationship points to a metathetic continuum of a substitutive system. To find support for a dual-process theory of motion perception Wagenaar and Vander Schaaf (1980), used continuously moving and stroboscopically presented stimuli, subjects that used magnitude estimation and category scaling techniques, and Stevens’ (1975) contention, as stated above, that additive and substitutive mechanisms yield different functional relationships. The results were inconclusive and, if anything, showed a tendency for a non-linear relationship for both stimulus conditions. The present experiment is a second attempt using the same scaling techniques, continuous and stroboscopic stimulus presentations, but with fewer velocities, twice as many judgments, and the presentation of the standard stimulus or range stimuli before every trial.
The experiment
Subjects The Ss were twelve male volunteers
from introductory, undergraduate classes. They had no previous experience in psychology experiments.
psychology
F. P. Vun Eyl /
Veloclt~ judgments
91
Apparatus The stimulus was a projected light dot, 3 cm in diameter (visual angle 24’), moving horizontally at eye level, from left to right, a distance of 170 cm (visual angle 21”). A stationary red dot, also 3 cm in diameter, serving as a fixation point, was 16 cm above the stimulus trajectory and in the S’s objective median plane. S was seated on a chair 457 cm from the screen on which the experimental and fixation stimuli were projected. Projection of the moving and fixation dots was in darkness and accomplished with two Kodak Carousel projectors. The white stimulus moved across the screen by way of a rotating mirror. Its travel distance was limited to 170 cm by a screen with a slit. Velocity of the moving stimulus was regulated with a rheostat. The stroboscopic effect was obtained by placing a rotating disc with a slit in the path of the light beam. The speed of the disc was chosen so that the stimulus was visible for 5 msec with a 5 Hz flicker frequency. Because stroboscopic light has less brightness than the same source presented continuously, the projection of the continuous stimulus was filtered (Kodak Wratten Filter No. 96, N.D. 0.10) to make the brightness of both conditions subjectively equal. Procedure Six Ss were used for the continuous light condition and six for the stroboscopic light condition. Under both conditions the stimulus was presented at speeds of 4, 8, 12, 16 and 20 deg/sec, four times, in random order. Half the Ss were instructed to use a magnitude scale first and then, after a IO-min rest, a category scale. The other half of the Ss used the two scaling techniques in opposite order. Scaling order was randomized over Ss. For both scaling techniques every stimulus presentation was preceeded by a standard medium speed (magnitude estimation), low speed stimulus and a high speed stimulus to indicate the range (category scale). In the case of magnitude estimation the standard was 100, for category scaling the fast stimulus was called 1 and the slow one 7. As the standard was presented E announced the value “100” and requested S’s estimate of the “next one”. During the category scaling condition E announced the values “1” and “7” or “7” and “1” which was also followed by a request for a value on the “next one”. During the experiment S was also instructed repeatedly to remain looking at the red fixation point and not to track the white dot. S was alerted to the appearance of each light by E saying “now”.
Table I
Mean judgments per rating scale, velocity condition, and light stimulus presentation. Velocity
deg/sec
Continuous light Stroboscopic light
Category
Magnitude
scale
estimation
20
16
12
8
4.
20
16
I2
8
4
6.8 6.6
6.3 5.6
4.6 4. I
2.8 3.2
1.5 1.3
135 139
124 I21
106 103
71 83
52 58
I
765-
I
1
. 0
P
1
I
I
. continuous o stroboscopic
:
L0 .
32-
. 0 I
I
I
L
8
12
velocity
I”
i
16 20
deg/sec
Fig. I. Velocity judgments of continuous (solid dot) and stroboscopic obtained with a category scaling technique.
(open dot) movement
as
Results All judgments tiere averaged over the Ss per continuous movement/category scaling condition, continuous movement/magnitude estimation condition, stroboscopic presentation/category scaling condition, and stroboscopic presentation/magnitude estimation condition. The results are presented in table 1 and graphically shown in figs. I-3. Figs. 1 and 2 show the judgments by category scale and magnitude estimation,
I
20( I-
continuous 7 - o stroboscopic l
15( )-
8
1O(I-
a
Is
z
.
ct c6 ,5 -t
1
.
a
5( l-
lo
0
5
a
: :
1
1
1
continuous o stroboscopic l
0
E 1
1 1
velocity
8 m
16 20
40 mean
deg/sec
Fig. 2. Velocity judgments of continuous obtained with magnitude estimation. Fig. 3. Magnitude estimations plotted against each other.
12
100 magnitude
(solid dot) and stroboscopic
and category
ratings
I
I
IO,
L
of continuous
(open
and
150 estlmatlon
dot) movements
stroboscopic
as
movement
F.P. Vun &vl /
Velocitv judgments
93
respectively. Fig. 3 has the category ratings plotted against those obtained with magnitude estimation. It is observed that differences between the continuous light data and stroboscopically obtained data are quite small and that all graphs have an unmistakably linear character. Of course, the logarithmic transformations in figs. 1 and 2 may contribute to the apparent linearity. Nevertheless, linear regression accounts for 78 percent of the variance of the mean category judgments and 76 percent of the mean magnitude estimation judgments for the continuous condition. Linear regression also accounts for 82 percent of the variance of the mean category judgments and 84 percent of the mean magnitude estimation judgments for the stroboscopic condition. The linear correlation between the magnitude estimation and the category scale means is r=0.80 for the continuous condition and r=0.78 for the stroboscopic condition. Since the data of both continuously and stroboscopically presented stimuli have linear continua, the dimensions are metathetic which leads to the conclusion that the mechanism or mechanisms by which they were processed are substitutive in nature.
Discussion In the present experiment perceived velocity following both continuous and stroboscopically presented stimuli was linearly related to physical velocity. This strongly suggests the possibility of a substitutive process for both dimensions. Wagenaar and Vander Schaaf (1980) on the other hand, found some inconclusive indications of non-linearity, suggesting the possibility of an additive process for both conditions. Although the present experiment limited itself to just one flicker frequency (5 Hz), this cannot be the source of discrepant findings because this frequency was the same one used by Wagenaar and Vander Schaaf (1980). A more likely source is the greater reliability of the present data because twice as many judgments were taken and the standard stimulus or the range was presented before every trial. Theoretically, the results of the present experiment do not support the possibility that continuous movement and movement as a result of abrupt, discontinuous changes of location differ because one adheres to a qualitative process, the other to a quantitative one. Unless Stevens’ (1975) rationale is faulty or does not apply to the present experiment, two possibilities suggest themselves: (1) there are two separate mechanisms but both are metathetic, or (2) there is just one mechanism and any dual-process notions should be abandoned. In so far as motion is defined as displacement in space, a truism about motion is that it always has direction. Quite conceivably velocity and direction are neuro-physiologically interrelated in such a way that information processed by velocity and direction detectors combine and interact at a higher level, If direction detection is contingent on changes in location, either because the background shifts against the target or the target shifts against the background, an uninterrupted stream of successive and small location changes is processed
more efficiently than a constantly interrupted flow of information about larger (more different) location changes. Particularly if it can be assumed that velocity detectors are non-functional during very short (5 msec) stimulus presentations in successive but different locations, it leaves any higher interpretive mechanism dependent on information of frequency and extent of location changes only. Of the two, frequency analysis is more likely to be the process that functions substitutively and, as such, the primary factor of inferred velocity during stroboscopic movement. Another possibility is that the primary factor is spatio-temporal modulation sensitivity and that the spatial and temporal separation ratios of stroboscopic presentations are processed substitutively. However, regardless of the outcome, answers can be expected to be within the context of processes that are substitutive rather than additive.
References Gregory, R.L.. 1964. Human perception. British Medical Bulletin, section 4: Perception of movement, 23-25. Lit. A., 1960. The magnitude of Pulfrich stereo phenomenon as a function of target velocity. Journal of Experimental Psychology 59, l65- 175. Richards, W.. 1971. Motion detection in man and other animals. Brain Behavior and Evolution 4. 162-181. Roufs, J.A., 1974. Dynamic properties of vision: V. Perception lag and reaction time in relation to flicker and flash thresholds. Vision Research 14 (9). X53-869. Sekuler, R.W.. A. Pantle and E. Levison, 1978. ‘Physiological basis of motion perception’. In: R. Held, H. Leibowitz and H.L. Teuber (eds.), Handbook of sensory physiology. VIIII. Berlin: Springer-Verlag. pp. 67-96. Stevens, S.S., 1975. Psychophysics (edited by G. Stevens). New York: Wiley. Stevens, S.S. and E.H. Galanter. 1957. Ratio scales and category scales for a dozen perceptual continua. Journal of Experimental Psychology 54, 377-41 I. Wagenaar, W.A. and T.W. vander Schaaf, 1980. SubJective anelheidsschalen van continue en stroboscopische stimuli. Report of the Institute for Perception, RVO-TNO. Soeaterbcrg. Netherlands. Wagenaar, W.A.. T.W. vander Schaaf and G.B. Flares d’Arcais. IYHOa. An experimental study of MacKay’s illusion: evidence for two movement perception systems. Report of the Institute for Perception, RVO-TNO, Soesterberg, Netherlands. Wagenaar, W.A.. S. Luthra and G.B. Florea d’Arcais, 1980b. Effects of luminance and velocity on perception lag in moving targets. Report of the Institute for Perception, RVO-TNO, Soestcrberg, Netherlands.