Article Figures & Data
Figures
- Figure 1.
Stimulus design and behavioral ratings. a, Rate × Pitch stimulus design. Stimuli were FM sounds that varied in modulation rate (decrease, no change, or increase over time) and pitch (decrease, no change, or increase over time). Here, the frequency modulation of example sounds is depicted, with the sounds' frequency displayed on the y-axis and time on the x-axis. Note that changes in modulation rate are schematic and were smaller in the actual stimuli. b, Mean behavioral ratings (±SEM) along the Rate and Pitch dimensions. Behavioral ratings are zero-centered, such that positive ratings belong to the percept of speeding up and negative values to the percept of slowing down.
- Figure 2.
Brain activity power spectra and source localizations. Top, Power spectra for the three pitch conditions at constant rate (mean over channels and participants). Asterisks mark significant enhancement of power at the sound's modulation rate (3.95 Hz; p < 0.001). Sensor topographical distributions are shown for the mean of the 3.9–4 Hz frequency bins. Bottom, Source localization of NAI at 3.95 Hz. Activity is displayed on a partially inflated standard brain surface, with dark gray representing sulci and light gray representing gyri. Enlarged displays of auditory cortices are provided for white dashed squares (LH, left hemisphere; RH, right hemisphere; HG, Heschl's gyrus; STG, superior temporal gyrus).
- Figure 3.
ITPC for the 3 Rate × 3 Pitch conditions. a, Sensor topographical distributions of ITPC values for each condition (mean across 1–4 s and 3.2–4.7 Hz). Bottom, 28 positions of the 56 channels over auditory cortices used for subsequent analyses. b, ITPC for all conditions (mean across the 56 channels over auditory cortices, see a). c, Mean estimated neural best frequency (±SEM) at sound offset (4.5 s) for each condition. d, Mean epoch-final ITPC strength (±SEM; 3.8–4 s time interval). e, Mean estimated slopes (of linear fits; ±SEM) reflecting change of ITPC strength over time.
- Figure 4.
Cerebro-acoustic phase lag. a, Topographical distributions of Bonferroni-corrected significance levels (PBonf) of the Pitch and the Rate main effects and the Rate × Pitch interactions from the circular ANOVA (Harrison–Kanji test). b, Individual phase lags (angles) and mean resultant vector for each condition as obtained at one gradiometer channel over right auditory cortex (marked by the black cycle in a).
- Figure 5.
Decoding of rate and pitch conditions from time-frequency phase and power patterns (1–4 s and 3.2–4.7 Hz). a, Decoding matrices showing proportions of classified trials with rows reflecting decoded labels and columns reflecting stimulus labels. The diagonal reflects decoding accuracies (proportion of correctly classified trials). b, Decoding accuracies from independent classifications within rate and pitch dimensions. Dashed lines indicate chance level and the asterisk marks decoding accuracies significantly larger than chance (*PBonf < 0.001).
- Figure 6.
Hemisphere-specific decoding accuracies and decoding of subjective ratings. a, Mean decoding accuracy (±SEM) for left and right auditory cortex regions using the phase data. b, Mean decoding accuracies (±SEM) from decoding the subjective perceptual ratings (slowing down; speeding up) from phase data and statistical results from testing decoding accuracies against chance level (dashed line). *PBonf < 0.05; #p < 0.10; n.s., not significant.
