BSCodec: A Band-Split Neural Codec for High-Quality Universal Audio Reconstruction

TL;DR

BSCodec is a neural audio codec that splits audio spectrum into bands for independent compression, improving high-quality reconstruction across speech, music, and sound while maintaining competitive speech quality.

Contribution

Introduces BSCodec, a novel neural audio codec architecture that processes spectral bands separately, addressing the limitations of speech-optimized codecs on diverse audio types.

Findings

Achieves superior reconstruction quality across diverse audio content.

Maintains competitive speech quality while improving music and sound fidelity.

Shows strong potential for downstream audio applications.

Abstract

Neural audio codecs have recently enabled high-fidelity reconstruction at high compression rates, especially for speech. However, speech and non-speech audio exhibit fundamentally different spectral characteristics: speech energy concentrates in narrow bands around pitch harmonics (80-400 Hz), while non-speech audio requires faithful reproduction across the full spectrum, particularly preserving higher frequencies that define timbre and texture. This poses a challenge: speech-optimized neural codecs suffer degradation on music or sound. Treating the full spectrum holistically is suboptimal: frequency bands have vastly different information density and perceptual importance by content type, yet full-band approaches apply uniform capacity across frequencies without accounting for these acoustic structures. To address this gap, we propose BSCodec (Band-Split Codec), a novel neural audio codec architecture that splits the spectral dimension into separate bands and compresses each band independently. Experimental results demonstrate that BSCodec achieves superior reconstruction over baselines across sound and music, while maintaining competitive quality in the speech domain, when trained on the same combined dataset of speech, music and sound. Downstream benchmark tasks further confirm that BSCodec shows strong potential for use in downstream applications.