Skip to main content

Real-time dynamic binaural synthesis with head tracking.

Project description



$ conda create --name binsim python=3.5 numpy
$ source activate binsim
$ pip install pybinsim

On linux, make sure that gcc and the development headers for libfftw and portaudio are installed, before invoking pip install pybinsim. For ubuntu:

$ apt-get install gcc portaudio19-dev libfftw3-dev


Create pyBinSimSettings.txt file with content like this

soundfile signals/test441kHz.wav
blockSize 512
filterSize 16384
filterList brirs/filter_list_kemar5.txt
maxChannels 2
samplingRate 44100
enableCrossfading True
useHeadphoneFilter False
loudnessFactor 0.5
loopSound False

Start Binaural Simulation

import pybinsim
import logging

pybinsim.logger.setLevel(logging.DEBUG)    # defaults to INFO
#Use logging.WARNING for printing warnings only

with pybinsim.BinSim('pyBinSimSettings.txt') as binsim:


Basic principle:

Depending on the number of input channels (wave-file channels) the corresponding number of virtual sound sources is created. The filter for each sound source can selected and activitated via OSC messages. The messages basically contain the number index of the source for which the filter should be switched and an identifier string to address the correct filter. The correspondence between parameter value and filter is determined by a filter list which can be adjusted individually for the specific use case.

Config parameter description:

Defines *.wav file which is played back at startup. Sound file can contain up to maxChannels audio channels. Also accepts multiple files separated by ‘#’; Example: ‘soundfile signals/sound1.wav#signals/sound2.wav
Number of samples which are processed per block. Low values reduce delay but increase cpu load.
Defines filter size of the filters loaded with the filter list. Filter size should be a mutltiple of blockSize.
Maximum number of sound sources/audio channels which can be controlled during runtime. The value for maxChannels must match or exceed the number of channels of soundFile(s).
Sample rate for filters and soundfiles. Caution: No automatic sample rate conversion.
Enable cross fade between audio blocks. Set ‘False’ or ‘True’.
Enables headhpone equalization. The filterset should contain a filter with the identifier HPFILTER. Set ‘False’ or ‘True’.
Factor for overall output loudness. Attention: Clipping may occur
Enables looping of sound file or sound file list. Set ‘False’ or ‘True’.

OSC Messages and filter lists:

Example line from filter list: 165 2 0 0 0 0 brirs/kemar5/kemar_0_165.wav

To activate this filter for the third channel (counting starts at zero) for your wav file you have to send the following message to the pc where pyBinSim runs (port 10000):

/pyBinSim 2 165 2 0 0 0 0

When you want to play another sound file you send:

/pyBinSimFile folder/file_new.wav

Or a sound file list

/pyBinSimFile folder/file_1.wav#folder/file_2.wav

The audiofile has to be located on the pc where pyBinSim runs. Files are not transmitted over network.


Check the repository for ready-to-use demos.


Please cite our work:

Neidhardt, A.; Klein, F.; Knoop, N. and Köllmer, T., “Flexible Python tool for dynamic binaural synthesis applications”, 142nd AES Convention, Berlin, 2017.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
pybinsim-1.2.1.tar.gz (17.2 kB) Copy SHA256 hash SHA256 Source None

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page