Surround.py: A Framework for Immersive Spatial Audio Simulation and Education

[ Link to Paper ]

Author: Varun Shenoy ([email protected])

Affiliation: Dept. of Electrical Engineering, Stanford University

John Carmack on HRTFs and Sound Localization

Project File Structure

• requirements.txt contains all the dependencies necessary to run this project. You can use conda or pip to install the exact versions of dependencies which we used from this file.
• The sounds folder contains various sample sounds from mixkit (https://mixkit.co/free-sound-effects/). You can add any sounds you want to try here, but make sure that they are 16 bit. You can use an online converter (https://audio.online-convert.com/convert-to-wav) to change a sound's bit resolution to 16 bit.
• The hrtf folder contains the entire KEMAR HRIR dataset from MIT. Information about this dataset can be found here: https://sound.media.mit.edu/resources/KEMAR.html.
• The code folder contains all the source files for the project.
  • Speakers.py houses the BasicSpeaker, HRTFSpeaker, and AmbisonicsSpeakers classes that are the bread and butter of this package.
  • Soundstage.py includes the Soundstage class and its various methods. The plot method is particularly comprehensive.
  • utils.py has several helper functions for interfacing with pydub, numpy, and the KEMAR dataset.
  • Take a look at the various demo files to see how this framework is used in action.

Running the project

I built this project on Python 3.8.11, but it might work on earlier versions.

After cd-ing into the main folder, install the prerequisite frameworks using pip install -r requirements.txt or your desired package manager.

Then, cd into the code folder and run python3 demo-hrtf.py or any other demo file that is already there (or one that you have created)!