THE AUDITORY MODELING TOOLBOX

Applies to version: 1.1.0

View the code

Go to function

REIJNIERS2014_FEATUREEXTRACTION - - extract HRTF using gammatone frequency bands and ITDs from SOFA object

Usage

[template, target] = reijniers2014_featureextraction(SOFAobj)

Input parameters

SOFAtemplate Struct in SOFA format with HRTFs

Output parameters

template template struct with spectral components
target template struct with spectral components

Description

reijniers2014_featureextraction(...) computes temporally integrated spectral magnitude profiles and itd.

reijniers2014_featureextraction accepts the following optional parameters:

'source_ir',source_ir
 Set the sound source's impulse reponse. Default value a broadband sound source with 0dB amplitude.
'fs',fs Set the sampling rate to fs. Default value takes the fs of SOFA template object.
'fb_ch',fb_ch Set the number of channels for the gammatone filterbank to fb_ch. Default value is 30.
'fb_low',fb_low Set the lowest frequency in the filterbank to fb_low. Default value is 300 Hz.
'fb_high',fb_high Set the highest frequency in the filterbank to fhigh. Default value is 15000 Hz.
'ir_pad',len Define the padding length for the impulse responses before being convolved with gammatone filters. Default value is 0.05 s.
'targ_az',targ_az Set the azimuth of a set of sound sources to targ_el. It can be a scalar or a column vector Default value is []: all target azimuths are used. Must have the same size of targ_el.
'targ_el',targ_el Set the elevation of a set of sound sources to targ_el. It can be a scalar or a column vector Default value is []: all target elevations are used. Must have the same size of targ_az.

References:

R. Barumerli, P. Majdak, R. Baumgartner, J. Reijniers, M. Geronazzo, and F. Avanzini. Predicting directional sound-localization of human listeners in both horizontal and vertical dimensions. In Audio Engineering Society Convention 148. Audio Engineering Society, 2020.

R. Barumerli, P. Majdak, R. Baumgartner, M. Geronazzo, and F. Avanzini. Evaluation of a human sound localization model based on bayesian inference. In Forum Acusticum, 2020.

J. Reijniers, D. Vanderleist, C. Jin, C. S., and H. Peremans. An ideal-observer model of human sound localization. Biological Cybernetics, 108:169--181, 2014.