Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Binamix -- A Python Library for Generating Binaural Audio Datasets

Dan Barry, Davoud Shariat Panah, Alessandro Ragano, Jan Skoglund, Andrew Hines

TL;DR

The paper addresses the need for large-scale, reproducible binaural audio datasets for spatial audio research, codec evaluation, and machine learning. It introduces Binamix, an open-source Python library that uses SADIE II HRIR/BRIR data to provide programmable binaural mixing, surround rendering, and IR interpolation. Key contributions include a TrackObject-based mixing architecture, a surround rendering pipeline with VBAP-based downmix, a modified Delaunay triangulation for IR interpolation, and a suite of example scripts and data loaders. The framework enables scalable, repeatable dataset generation across diverse speaker layouts and conditions, supporting reproducible research in spatial audio coding, quality metrics, and machine learning tasks.

Abstract

The increasing demand for spatial audio in applications such as virtual reality, immersive media, and spatial audio research necessitates robust solutions to generate binaural audio data sets for use in testing and validation. Binamix is an open-source Python library designed to facilitate programmatic binaural mixing using the extensive SADIE II Database, which provides Head Related Impulse Response (HRIR) and Binaural Room Impulse Response (BRIR) data for 20 subjects. The Binamix library provides a flexible and repeatable framework for creating large-scale spatial audio datasets, making it an invaluable resource for codec evaluation, audio quality metric development, and machine learning model training. A range of pre-built example scripts, utility functions, and visualization plots further streamline the process of custom pipeline creation. This paper presents an overview of the library's capabilities, including binaural rendering, impulse response interpolation, and multi-track mixing for various speaker layouts. The tools utilize a modified Delaunay triangulation technique to achieve accurate HRIR/BRIR interpolation where desired angles are not present in the data. By supporting a wide range of parameters such as azimuth, elevation, subject Impulse Responses (IRs), speaker layouts, mixing controls, and more, the library enables researchers to create large binaural datasets for any downstream purpose. Binamix empowers researchers and developers to advance spatial audio applications with reproducible methodologies by offering an open-source solution for binaural rendering and dataset generation. We release the library under the Apache 2.0 License at https://github.com/QxLabIreland/Binamix/

Binamix -- A Python Library for Generating Binaural Audio Datasets

TL;DR

The paper addresses the need for large-scale, reproducible binaural audio datasets for spatial audio research, codec evaluation, and machine learning. It introduces Binamix, an open-source Python library that uses SADIE II HRIR/BRIR data to provide programmable binaural mixing, surround rendering, and IR interpolation. Key contributions include a TrackObject-based mixing architecture, a surround rendering pipeline with VBAP-based downmix, a modified Delaunay triangulation for IR interpolation, and a suite of example scripts and data loaders. The framework enables scalable, repeatable dataset generation across diverse speaker layouts and conditions, supporting reproducible research in spatial audio coding, quality metrics, and machine learning tasks.

Abstract

The increasing demand for spatial audio in applications such as virtual reality, immersive media, and spatial audio research necessitates robust solutions to generate binaural audio data sets for use in testing and validation. Binamix is an open-source Python library designed to facilitate programmatic binaural mixing using the extensive SADIE II Database, which provides Head Related Impulse Response (HRIR) and Binaural Room Impulse Response (BRIR) data for 20 subjects. The Binamix library provides a flexible and repeatable framework for creating large-scale spatial audio datasets, making it an invaluable resource for codec evaluation, audio quality metric development, and machine learning model training. A range of pre-built example scripts, utility functions, and visualization plots further streamline the process of custom pipeline creation. This paper presents an overview of the library's capabilities, including binaural rendering, impulse response interpolation, and multi-track mixing for various speaker layouts. The tools utilize a modified Delaunay triangulation technique to achieve accurate HRIR/BRIR interpolation where desired angles are not present in the data. By supporting a wide range of parameters such as azimuth, elevation, subject Impulse Responses (IRs), speaker layouts, mixing controls, and more, the library enables researchers to create large binaural datasets for any downstream purpose. Binamix empowers researchers and developers to advance spatial audio applications with reproducible methodologies by offering an open-source solution for binaural rendering and dataset generation. We release the library under the Apache 2.0 License at https://github.com/QxLabIreland/Binamix/
Paper Structure (14 sections, 6 figures)

This paper contains 14 sections, 6 figures.

Table of Contents

  1. Introduction
  2. Binamix
  3. Motivation
  4. Tools Overview
  5. Applications
  6. Method and Design
  7. Mixing Architecture
  8. Surround Rendering and Layout Simulation
  9. Interpolation Method
  10. Delaunay Triangulation
  11. Interpolation Modes
  12. Example Scripts
  13. Conclusion
  14. Acknowledgments

Figures (6)

  • Figure 1: Binamix system overview. Blocks in green are currently supported or *coming soon.
  • Figure 2: Binamix signal flow.
  • Figure 3: Example of binaural mixing
  • Figure 4: Example of channel-encoded surround rendering
  • Figure 5: Delaunay Triangulation on two spatial distributions: 50-point Lebedev (left) and 7.1.4 layout (right).
  • ...and 1 more figures