Bendik Christensen,
Version 1 of Dataset published 2024 via 4TU.ResearchData
This project involves conducting and analyzing listening tests using modified webMUSHRA software to evaluate the perceptual accuracy of simulated acoustic environments. The code is structured into five main directories: Docs, containing ethics documents; Modified webMUSHRA Software, including testing code and configurations run with Docker for paired_comparison and subjective_eval tests; Results, storing both raw and processed data from the listening tests; Samples, providing original and convolved audio files with real and simulated Room Impulse Responses (RIRs); and Utils, featuring scripts for generating sine sweeps, convolving and clipping audio, and performing basic statistical analysis. For questions, contact b.christensen@student.tudelft.nl.