Skip to main content
. 2010 Nov 24;30(47):15969–15980. doi: 10.1523/JNEUROSCI.0966-10.2010

Figure 8.

Figure 8.

CNIC filtering characteristics enhance encoding efficiency for natural sound ensembles. A, Temporal ensemble efficiency for an equal resolution filterbank (gray) and a proportional resolution filterbank where filter bandwidths scale as for CNIC (black). The CNIC filterbank exhibits significantly higher temporal efficiency (p < 0.01, bootstrap t test) for all three natural sound ensembles tested (SP, speech; VOC, animal vocalizations; BACK, background; ALL, SP+VOC+BACK). B, Spectral encoding efficiency is shown for the equal resolution filterbank (gray) and CNIC model filterbank (black). The spectral ensemble efficiency is significantly higher for the CNIC model filterbank for all natural sound ensembles tested (p < 0.01, bootstrap t test). Error bars designate SEM.