Figure 7. The inhibitory tuning latencies and reverberation times show similar frequency dependence.

(A) Center of mass values ($COM$) are plotted against the anechoic frequency channel being estimated, for the excitatory and inhibitory fields of each model kernel for the large room and for the small room. These are color coded as follows: excitatory $COM$ (large room, $CO{M}_{large}^{+}$, red; small room, $CO{M}_{small}^{+}$, pink) and their inhibitory counterparts ($CO{M}_{large}^{-}$, blue; $CO{M}_{small}^{-}$, cyan). The dashed lines show a linear regression fit for each room, and the Pearson’s r value for each fit is given at the top of each the plot. (B) $COM$ values are plotted against the best frequency for the neuronal data (sound frequency of highest STRF weight). Each cortical unit was assigned a best frequency and the $COM$ values measured. The solid lines represent the mean $COM$ value for each best frequency, the shaded areas show ± SEM; color scheme and other aspects as in A. (C) $R{T}_{60}$ and $R{T}_{10}$ values are plotted as a function of cochlear frequency bands, for the large (dark green) and small (light green) rooms. Linear regression fit (dotted line) was used as in A and B to calculate r. Significance of Pearson’s correlation: ${\displaystyle {}^{\ast \ast \ast \ast }p<0.0001,{}^{\ast \ast }p<0.01}$.

Figure 7—figure supplement 1. Binaural room impulse responses.

Spectrograms of the binaural room impulse responses are plotted. (A) The left panel shows the left ear impulse response of the small room, while the right panel shows that of the right ear. (B) Same as A, but the spectrogram of the left and right ear of the large room impulse responses are shown. In all panels in A and B, the gray scale represents the sound energy in decibels (dB). The spectrograms were created using 5ms windows with 2.5ms overlap.