On the Geometry of Music Bandwidth Extension in Latent Spaces of Audio Codecs

2026-08-04

AI Summary

This work addresses the degradation in audio quality caused by bandwidth limitation in music signals, a problem typically tackled by complex generative models. The authors observe that the latent spaces of certain neural audio codecs inherently align with audio bandwidth characteristics. Leveraging this insight, they propose an efficient bandwidth extension method based on the geometric structure of the latent space: by estimating a transfer vector between clean and degraded latent representations on a reference set, they directly apply this vector as an additive correction to degraded latents. Despite its simplicity—relying solely on arithmetic operations in the latent domain—the approach achieves performance comparable to state-of-the-art diffusion-based models across multiple codecs. These results validate the efficacy of latent-space geometry for audio restoration and offer a new, computationally efficient paradigm for high-quality audio enhancement.

Abstract

Recent audio restoration increasingly relies on large-scale conditional latent generative modeling, including diffusion, Schrodinger Bridges, and Flow Matching variants, to invert degradations such as bandwidth limitation or noise. We present an analysis of the performance of various state-of-the-art methods compared to simple arithmetic transformations in the latent spaces of multiple neural codecs for musical bandwidth extension. We show that estimating a single transport vector between the clean and degraded latent centroids on a reference set, and adding it to degraded latents, can yield restoration performance competitive with large diffusion models. This suggests, first, that some neural codec latent spaces exhibit structure aligned with audio bandwidth; and second, that in such cases complex conditional models may offer only limited gains over a simple vector addition. We argue that these findings reveal an interesting avenue for future research whereby models could take advantage of the latent space structure in order to offer greater training and parameter efficiency, and overall better performance. Additionally, we propose to consider this simple arithmetic transformation as a baseline for music bandwidth extension research, as it allows an assessment of the contribution of learnable parameters towards restoration performance.

Problem

• music bandwidth extension
• latent space
• audio codecs
• audio restoration
• bandwidth limitation

Innovation

• latent space geometry
• bandwidth extension
• neural audio codecs
• vector arithmetic
• audio restoration