Home » Software » Spatialization by equalizing filters

Spatialization by equalizing filters

These Matlab routines are used to create spatializations via inverse filters and loudspeaker grouping in the way is explained in [1].

To use them, please open the file “eqf_spatialization.m” and edit it according to your needs (destination path, audio to spatialize, etc.). These routines are meant to be used with a diffuse-field equalized version of the database provided by Qu et al. [2]. The original database can be found here:


Note that this database is copyrighted by the Key Lab of Machine Perception (Ministry of Education), Peking University. It is provided free with no restrictions on research use, provided the authors are cited when the data is used. And it is forbidden to be used in any commercial application.

The code was developed with Matlab Version (R2016b), it works with current versions of Matlab ( (R2018b) Prerelease) and probably with some earlier versions as well. It uses the Signal Processing Toolbox.

To download, please use the bitbucket repository:

Bitbucket repository for EQF


In this implementation, users may edit the main script to indicate the directory path of a monophonic sound source, the location of the HRIR database, the destination path of the spatialized sounds, the actual location of loudspeakers, and the location of the sound source. Spatialized sounds are stored as Extensible Wave-Format files (six-channel format, corresponding to Left, Right, Center, LF, Surround Left, and Surround Right loudspeakers).


To better experience while listening to the following examples, please make sure that you have a loudspeaker layout of five loudspeakers located at a radius of 1.4 m, at the same height of the listener’s ears, and at azimuths [0°, ±30°, ±120°]. A player based in Pure-data is also offered in case of need.

6chPlayer (a player for files in format WAVE_EXTENDED)

Spatialized Helicopter (The sound of a helicopter simulated to move from the front to the zenith, then to the right, again to the zenith, then to the back, around to the left, and finally moving to the back).

Spatialized Drumset (A drumset riff spatialized at different azimuths and elevations, presented sequentially).


[1] J. Villegas, “Elevation perception improvement in loudspeakers rings by means of equalizing filters,” Acoust. Sci. & Tech., 2018.
[2] T. Qu, Z. Xiao, M. Gong, Y. Huang, X. Li, and X. Wu, “Distance-Dependent Head-Related Transfer Functions Measured With High Spatial Resolution Using a Spark Gap,” IEEE Trans. on Audio, Speech & Language Processing, vol. 17, no. 6, pp. 1124–1132, 2009.