# Audio Similarity is Unreliable as a Proxy for Audio Quality

arXiv: https://arxiv.org/abs/2206.13411 · HTML: https://mlsanity.com/paper/2206.13411/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Jun 2022.

## Built on by (most cited first, 7 of 7)

- [Survey on the Evaluation of Generative Models in Music](https://mlsanity.com/paper/2506.05104/) (2025)
- [FINALLY: fast and universal speech enhancement with studio-like quality](https://mlsanity.com/paper/2410.05920/) (2024)
- [VoiceBridge: General Speech Restoration with One-step Latent Bridge Models](https://mlsanity.com/paper/2509.25275/) (2025)
- [Arab Voices: Mapping Standard and Dialectal Arabic Speech Technology](https://mlsanity.com/paper/2601.13319/) (2026)
- [CORN: Co-Trained Full- And No-Reference Speech Quality Assessment](https://mlsanity.com/paper/2310.09388/) (2023)
- [JSQA: Speech Quality Assessment with Perceptually-Inspired Contrastive Pretraining Based on JND Audio Pairs](https://mlsanity.com/paper/2507.11636/) (2025)
- [UniPASE: A Generative Model for Universal Speech Enhancement with High Fidelity and Low Hallucinations](https://mlsanity.com/paper/2604.14606/) (2026)

## Abstract (arXiv)

Many audio processing tasks require perceptual assessment. However, the time and expense of obtaining ``gold standard'' human judgments limit the availability of such data. Most applications incorporate full reference or other similarity-based metrics (e.g. PESQ) that depend on a clean reference. Researchers have relied on such metrics to evaluate and compare various proposed methods, often concluding that small, measured differences imply one is more effective than another. This paper demonstrates several practical scenarios where similarity metrics fail to agree with human perception, because they: (1) vary with clean references; (2) rely on attributes that humans factor out when considering quality, and (3) are sensitive to imperceptible signal level differences. In those scenarios, we show that no-reference metrics do not suffer from such shortcomings and correlate better with human perception. We conclude therefore that similarity serves as an unreliable proxy for audio quality.