Models of population-level frequency tuning sharpness. Schematic models of population-level frequency tuning sharpness with respect to the constant sequencing (test stimulus (TS) with constant frequency; left column) and the random sequencing conditions (random TS frequency; right column). The arrows indicate the presented frequencies. 'Gain' size is represented as neural activity amplitude; degree of 'sharpening' corresponds to the width of the frequency band that effectively evokes neural activity. A: Neural activity corresponding to the TS frequency (thick line) and other frequencies (thin line). In the constant sequencing condition, the neural activity corresponding to the constant TS frequency is larger and frequency tuning is sharper compared to other frequencies, due to frequency-specific 'gain' and 'sharpening' effects. In contrast, in the random-sequencing condition, 'gain' and 'sharpening' effects are widely distributed across frequencies, resulting in identical frequency tuning sharpness for TS and non-TS frequencies. B: Neural activity elicited by TS and band-eliminated noises (BENs). Left and right columns represent the constant sequencing (Constant TS) and random sequencing (Random TS) conditions. The top, middle, and bottom rows represent wide (1 critical band (CB)), middle (1/2 CB), and narrow (1/4 CB) BEN conditions. The three differently colored areas represent three distinct neural groups: (i) neurons merely activated by BEN (light gray areas), (ii) neurons merely activated by TS (dark gray areas), and (iii) neurons activated by both BEN and TS (black areas). The dark gray areas correspond to N1m source strength elicited by TS-onset, since the neural activity represented by the light gray and black areas has been masked by the simultaneously presented (and earlier onsetting) BEN. Notably, the neural activity surrounding the TS frequency in the constant sequencing condition is larger and sharper due to the frequency-specific 'gain' and 'sharpening' effects, as shown in Figure 1A.