Interaural time differences (itdProc.m)

The ITD between the left and the right ear signal is estimated for individual frequency channels and time frames by locating the time lag that corresponds to the most prominent peak in the normalised CCF. This estimation is further refined by a parabolic interpolation stage [May2011], [May2013b]. The ITD processor does not have any adjustable parameters, but it relies on the CCF described in Cross-correlation (crosscorrelationProc.m) and its corresponding parameters (see Table 32). The ITD representation is computed by using the request entry ’itd’.

The ITD processor is demonstrated by the script DEMO_ITD.m, which produces two plots as shown in Fig. 38. The ear signals for a speech source that is located closer to the right ear are shown in the left panel. The corresponding ITD estimation is presented for each individual TF unit (right panel). Apart from a few estimation errors, the estimated ITD between both ears is in the range of 0.5 ms for the majority of TF units.


Fig. 38 Binaural speech signal (left panel) and the estimated ITD in ms shown as a function of time frames and frequency channels.

[May2011]May, T., van de Par, S., and Kohlrausch, A. (2011), “A probabilistic model for robust localization based on a binaural auditory front-end,” IEEE Transactions on Audio, Speech, and Language Processing 19(1), pp. 1–13.
[May2013b]May, T., van de Par, S., and Kohlrausch, A. (2013), “Binaural Localization and Detection of Speakers in Complex Acoustic Scenes,” in The technology of binaural listening, edited by J. Blauert, Springer, Berlin–Heidelberg–New York NY, chap. 15, pp. 397–425.