Immersive Formats Evaluation
Reproducible objective benchmarking of scene-based, object-based, and binaural audio formats.
Supplementary research repository with raw model outputs, reproducible R analysis scripts, and report artefacts for an objective comparison of scene-based (Higher-Order Ambisonics), object-based (Dolby Atmos), and binaural immersive audio formats. Quality assessment uses the BAM-Q and BINAQUAL binaural quality models applied to a choral HOA production as the reference signal.
A key finding is that Ambisonics formats discriminate perceptual differences 3.79× more than channel-based formats under the tested paradigm. The repository supports the companion paper submitted to Applied Acoustics (2026). Designed for transparent methodology and repeatable benchmarking workflows.