Skip to content
/ KSRIR Public

Official code for KSRIR – an objective evaluation metric for B-format RIRs.

License

Notifications You must be signed in to change notification settings

HualinR/KSRIR

Repository files navigation

🎧 A Metric for Predicting the Quality of Ambisonic Spatial Audio Reproduced Using Spatially Interpolated or Extrapolated Room Impulse Responses

This repository contains the official implementation of the paper:

A Metric for Predicting the Quality of Ambisonic Spatial Audio Reproduced Using Spatially Interpolated or Extrapolated Room Impulse Responses,
presented at ICASSP 2025.

This work proposes a novel objective evaluation metric for assessing the quality of B-format Room Impulse Responses (RIRs). These RIRs are commonly used in ambisonic spatial audio systems, particularly in applications involving spatial interpolation or extrapolation of sound fields.


📦 Repository Overview

  • ├── ICASSP25_test.m – Main script to run the evaluation
  • ├── ICASSP25_KSRIR.m – Core objective quality metric
  • ├── xRIRAnalyzer_HR.m – Segment-level analysis with configurable parameters
  • ├── xRIRCleaner.m – Preprocessing: removes low-level and silent parts before the direct sound
  • ├── getWindowCenteredAt.m – Utility function for segment windowing
  • ├── getShortTermAverage.m – Helper for local averaging
  • └── *.mat – Supporting reflection profile data

🚀 Getting Started

Prerequisites

  • MATLAB (tested on R2021b or later)
  • Reference and synthesized WAV-format ambisonic audio files (Ensure both are at the same sampling rate and have equal length)

Running the Code

  1. Clone or download this repository.
  2. Open ICASSP25_test.m in MATLAB.
  3. Set the path to your own WAV files inside the script.
  4. Run the script to start the evaluation.

⚙️ Configuration

Adjust Segment Length: Modify the segment length parameter tau3 in xRIRAnalyzer_HR.m, to change the temporal resolution of the analysis.


📈 Output

Running the evaluation will output an objective spatial audio quality score based on the proposed metric. This score is useful for:

Benchmarking spatial interpolation or extrapolation algorithms

Evaluating room acoustics processing

Supporting perceptual quality assessment studies in immersive audio


📄 Citation

If you use this code or reference the method in your work, please cite the following:

@INPROCEEDINGS{10889495,
  author    = {Ren, Hualin and Ritz, Christian and Zhao, Jiahong and Zheng, Xiguang and Jang, Daeyoung},
  booktitle = {ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, 
  title     = {A Metric for Predicting the Quality of Ambisonic Spatial Audio Reproduced Using Spatially Interpolated or Extrapolated Room Impulse Responses}, 
  year      = {2025},
  volume    = {},
  number    = {},
  pages     = {1-5},
  doi       = {10.1109/ICASSP49660.2025.10889495}
}

If you use components related to direct sound and reflection capture or segment analysis (xRIRAnalyzer_HR.m, xRIRCleaner.m, getWindowCenteredAt.m, getShortTermAverage.m, or *.mat), please also cite:

@Article{app12042061,
  author    = {Zhao, Jiahong and Zheng, Xiguang and Ritz, Christian and Jang, Daeyoung},
  title     = {Interpolating the Directional Room Impulse Response for Dynamic Spatial Audio Reproduction},
  journal   = {Applied Sciences},
  year      = {2022},
  volume    = {12},
  number    = {4},
  article-number = {2061},
  url       = {https://www.mdpi.com/2076-3417/12/4/2061}
}

📬 Contact

For questions, feedback, or collaboration, feel free to contact:

📧 Hualin Ren — hualin@uow.edu.au


📝 License

This repository is made available for academic and research use only.

For commercial licensing, please contact the authors directly.

About

Official code for KSRIR – an objective evaluation metric for B-format RIRs.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages