14.2 Subjective Preferences for Oscillatory Movements

sentations have been traditionally based on recognizing spectral patterns; however, pitch of complex tones and timbre, for example, can be well described by temporal factors extracted from the ACF. It is worth noting that sensory information from other modalities besides audition and vision can be handled using similar autocorrelation and crosscorrelation neurocomputational frameworks (Cariani, 2001).

14.2 Subjective Preferences for Oscillatory Movements

Preference judgments using the PCT for sinusoidal movements of a single circular target without any fluctuation on a monitor screen were performed. The period of stimulus movements was varied separately in the vertical or horizontal direction. Results show that the most preferred periods ([T]p) for all subjects are about 1 s in the vertical direction and about 1.3 s in the horizontal direction. The curve of the scale values of preference may be commonly expressed by Equation (14.1) with x = log10T – log10[T]p and β = 3/2. All observers participating in the vertical direction series showed that the curves for the scale value of preference are significantly steeper in the fast-moving range in reference to [T]p than those in the slow-moving range.

It has been shown as an auditory sensation that the preferred repetition period for noise bursts was approximately 0.55 s. In vision, subjective preference for a flickering light showed that the preferred period was approximately double this (Soeta et al., 2002a; Soeta et al., 2002c). As is described in Section 3.3, results of the scale value of subjective preference from the different test series, using different music programs, yield the common formula of Equation (3.9) as well as Equation (14.1).

Ten subjects (21–26 years of age) participated. All subjects had normal or correct-to-normal binocular vision. The stimuli were displayed on a CRT monitor presenting 30 frames per second. Figure 14.6 shows the stimulus, a single, white, circular target moving sinusoidal, used in the experiment. The diameter of the target was subtended 1◦ of the visual angle (1.22 cm). The movement of the stimulus is expressed as

where A is the amplitude and T is the period of the stimulus. In all experiments, the amplitude A was fixed at 0.61 cm on the monitor screen, corresponding to 0.5◦ of visual angle. The white target and black background corresponded with gray levels 40 and 0.5 cd/m2, respectively. The monitor presenting the stimuli was placed in a dark room 0.7 m away from the subject’s eye position to maintain natural binocular vision.

Subjective preference for the period of movements in the horizontal and vertical directions was examined separately. The period of stimulus movement T in Equation (14.2) was varied at six levels: T = 0.6, 0.8, 1.2, 1.6, 2.0, and 2.4 s. Thirty

Fig. 14.6 Stimulus target used in the experiment showing an example of oscillatory horizontal movement

pairs combining six different periods constituted each series, and 10 series were conducted for all 10 subjects in the experiments by the PCT.

The most preferred period [T]p for each subject was estimated by fitting a suitable polynomial curve to a graph on which scale values were plotted. Figure 14.7 shows an example of the method used for estimating [T]p. The peak of this curve denotes the subject’s most preferred value. Table 14.4 shows results of the most preferred periods for each subject for both vertical and horizontal directional stimuli. The global value of the most preferred period was about 0.97 s for vertical movement and about 1.26 s for horizontal movement. Results from all subjects indicated that preferred periods in the vertical direction were shorter than that of those in the horizontal direction (p < 0.01).

Fig. 14.7 Viewer preferences for rate of visual oscillatory motion. The arrow indicates the most preferred period [T]p (≈ 1.10 [s]) in the vertical direction (subject J)

We also attempted to determine the characteristics of the preference evaluation curve in more detail. As shown in Fig. 14.7, the preference evaluation curve here can also be expressed in the form of Equation (14.1), where x is replaced by log10T

– log10[T]p. After obtaining the most preferred period for each subject, we identified values of α and β for the period in the fast-moving range in reference to [T]p, and also in the slow-moving range (Table 14.5). The values of α and β in the fastmoving range in reference to [T]p for the vertical direction could not be calculated, because there are only two available scale values. The average value of α, estimated by a quasi-Newton numerical method, was approximately 1.43. Thus, the value of α can be fixed at 3/2 similar to that above. The weighting coefficient α describes the sharpness of the preference curve with respect to the normalized period. Values of α of each individual for the period of T < [T]p and T > [T]p may be obtained as indicated in Table 14.6.

Table 14.5 The values of α and β for each subject as calculated by Equation (14.1) and averaged values of β. Because averaged values of β are close to 3/2, it may be fixed at 3/2 obtaining a single constant α representing the individual difference as listed in Table 14.6

Table 14.6 The values of α obtained in the ranges of fast and slow periods with respect to the most preferred period [T]p: αf(≤[T]p) and αs(≥[T]p) for each subject. When the value of β is fixed at 3/2 in Equation (14.1), then the individual differences may be represented by the constant α

Fig. 14.8 (a) The normalized scale values of preference for individual subjects for movement oscillation period in the vertical direction. (b) Those in the horizontal direction. Different symbols indicate scale values obtained with different subjects

Figure 14.8 shows scale values for all subjects and the preference evaluation curve calculated by Equation (14.1). The results indicate that a sinusoidal period of about 1.26 s is preferred for horizontally moving stimuli. This period is approximately twice the period of the most preferred tempo for noise bursts and the same as the most preferred period of a flickering light. For vertically moving stimuli, the most preferred period was about 0.97 s, which is a significantly faster period than that of the horizontally moving stimuli. Moreover, the values of α in the range of T < [T]p were significantly larger than those of T > [T]p in regard to vertical movement (p < 0.01).

Several investigators found that motion sensitivity to vertical and horizontal movement is isotropic (Ball and Sekuler, 1979; Levinson and Sekuler, 1980; van de Grind et al., 1993; Raymond, 1994; Gros et al., 1998). Kinchla and Allan (1970)