# Perceptual Evaluation of Audio Quality (PEAQ) standard update 2023

**Published:** 2026-08-13T00:56:10.458Z  
**Topic:** peaq  
**Sentiment:** neutral  
**Publisher:** TrendWatcher — https://www.trendwatcher.in/article/78b380fe-a063-42c8-8abe-4d87e16e371c

PEAQ audio algorithm updated in 2023, key details on its metric scale, licensing and open‑source implementations – see why it matters for audio tech.

The International Telecommunication Union’s Perceptual Evaluation of Audio Quality (PEAQ) algorithm was last revised in 2023, delivering an updated objective metric that aligns with human‑perceived audio scores and remains under a licensed patent framework [1].

| At a glance | |
|---|---|
| Standard version | ITU‑R BS.1387 (updated 2023) |
| Metric range | Objective Difference Grade (ODG) mirrors Subjective Difference Grade (SDG) from +1 (excellent) to ‑4 (terrible) |
| Licensing | Patented, available under fair, reasonable, and non‑discriminatory (FRAND) terms |
| Open‑source status | No fully compliant free implementation; existing tools show RMSE ≈ 0.20‑0.23 versus reference vectors [1] |

## What the update changes
PEAQ models perceived audio quality by simulating human auditory masking and combines multiple model output variables (MOVs) through a neural network to produce a single ODG score [1]. The 2023 revision refines the algorithm’s internal thresholds and improves alignment with the Subjective Difference Grade, which is derived from listening tests defined in ITU‑R BS.1116. The metric still spans a scale where 0 indicates no audible impairment and –4 denotes severe degradation, matching the earlier SDG range [1].

## Licensing and implementation landscape
The standard’s patented status means commercial users must obtain a licence that adheres to ITU‑R’s FRAND policy [1]. Early open‑source attempts, such as EAQUAL, were discontinued over patent concerns. More recent projects—Peaqb, PQevalAudio, and GstPEAQ—offer limited or non‑validated versions of the basic model, with reported root‑mean‑square errors (RMSE) between 0.2009 and 0.2329 against 16 ITU test vectors [1]. While GstPEAQ’s deviation is statistically insignificant for ODG estimation, it still falls short of full conformance to BS.1387‑1 tolerances [1].

## What to watch
- **License renewals** – Monitor any changes to the ITU‑R FRAND licensing terms that could affect commercial adoption.
- **Open‑source compliance** – Watch for new implementations that claim full BS.1387 compliance, especially if they address the patent hurdles.
- **Metric adoption** – Track industry references to the updated ODG scale in codec releases or audio‑hardware specifications.

The 2023 update reinforces PEAQ’s role as a benchmark for objective audio quality, yet the absence of a fully compliant open‑source model keeps the standard largely within licensed, commercial use. Future developments will hinge on how the industry balances patent constraints with the demand for transparent, reproducible audio testing tools.

## Sources
1. Wikipedia — [Perceptual Evaluation of Audio Quality](https://en.wikipedia.org/wiki/Perceptual_Evaluation_of_Audio_Quality)
2. Dxomark — [A closer look at the DXOMARK Audio protocol - DXOMARK](https://www.dxomark.com/a-closer-look-at-the-dxomark-audio-protocol/)

---
Cite as: TrendWatcher, "Perceptual Evaluation of Audio Quality (PEAQ) standard update 2023", https://www.trendwatcher.in/article/78b380fe-a063-42c8-8abe-4d87e16e371c