Figure 1.
Block diagram of the frequency-variable sound presentation system.
Each signal of the stereophonic sound material was divided into a low frequency component (LFC) and a high frequency component (HFC). The LFC was low-pass filtered at the cut off frequency of 16 kHz, while the HFC was band-pass filtered at arbitrary frequencies over 16 kHz by a programmable band-pass filter. The LFC and the HFC were independently amplified and reproduced through speakers and super-tweeters. Either LFC alone (control) or the LFC and an HFC together was applied to the subjects to observe the emerging state of the hypersonic effect.
Figure 2.
Averaged power spectra of sound materials applied in Experiment 1.
A: Electric signal of the sound source. B: Filtered electric signals of LFC under 16 kHz (control), HFC at 16–48 kHz, and HFC at 48 kHz <. C: The sounds and air vibration reproduced by the frequency-variable sound application system. Power was calculated from the data recorded at the subject's position. Faithful reproductions from electric signals to air vibrations were observed.
Figure 3.
Averaged power spectra of sound materials applied in Experiment 2.
Filtered electric signals of LFC under 16(control) and HFCs above 16 kHz. Frequency from 16 kHz to 96 kHz was divided into ten bandwidth components at every 8 kHz, while higher frequencies above 96 kHz were divided into two bandwidths, one at 96 kHz–112 kHz, the other at 112 kHz <. A filtered HFC was applied together with an LFC to the subjects before or after LFC alone (control) was applied.
Figure 4.
Results of Experiment 1: broad examination of frequency dependency of the hypersonic effect.
A: Scalp distribution of electroencephalographic activity during application of different HFC frequencies. To overview the spatial distribution of the equivalent potential of alpha2 frequency band (10–13 Hz) of EEGs, colored contour line maps were constructed by using 2,565 scalp grid points computed by linear interpolation and extrapolation of alpha2 components from 12 electrodes [33], [34]. Darker red indicates higher alpha2. Note that the alpha2 in the occipital region changed depending on the frequency of the HFC. B: Mean (+SE) value of Alpha-2 EEGs. Potential of alpha2 frequency band recorded from 7 electrodes in the centro-parieto-occipital region (C3, C4, T5, Pz, T6, O1 and O2) for the last 100 sec of sound application was averaged across all subjects. Analysis of variance (ANOVA) revealed the main effect of HFC to be significant, and Tukey's post-hoc test found significant difference between [LFC+HFC16-48] and [LFC+HFC48<].
Figure 5.
Results of Experiment 2: detailed examinations of frequency dependency of the hypersonic effect.
A: Mean (+SE) values of ΔAlpha-2 EEG across the subjects in each of twelve sub-experiments. The frequencies indicate the frequency range of HFC that were applied together with LFC ([LFC+HFC]) in comparison with the control ([LFC] alone). ΔAlpha-2 EEG is calculated by subtracting Alpha-2 EEG obtained during [LFC] from those during [LFC+HFC] in each subject. Univariate ANOVA showed the main effect of the frequency of HFC (p<0.05). Tukey's post-hoc tests showed a significant difference between ΔAlpha-2 EEG obtained with [LFC+HFC16-24] and that obtained with [LFC+HFC80–88] (p<0.05). B: Comparison of ΔAlpha-2 EEG between two groups of pooled data obtained in the sub-experiments using HFC below 32 kHz and those obtained in the sub-experiments using HFC above 32 kHz. Unpaired t-tests showed significant difference between the two groups and 1-sampled t-tests showed significant difference from zero for each group, that is, ΔAlpha-2 EEG obtained by using HFC below 32 kHz was significantly negative (p<0.01), while those obtained by using HFC above 32 kHz was significantly positive (p<0.001).