Metrics › Loudness

Time-varying loudness, Moore–Glasberg–Schlittenlacher (ISO 532-3)

Unit: sone, phon · Also known as ISO 532-3, TVL, time-varying loudness, Glasberg and Moore 2002, Moore-Glasberg-Schlittenlacher method, short-term loudness, long-term loudness

Loudness of time-varying sounds as instantaneous, short-term and long-term loudness. It starts from the Glasberg & Moore (2002) model; Moore et al. (2016) added binaural inhibition and Moore et al. (2018) revised the time constants. ISO 532-3:2023 standardises the resulting Moore–Glasberg–Schlittenlacher method for stationary and time-varying sounds.

Editions and who implements them

Project map

The project map, with only the lines for time-varying loudness (Moore–Glasberg–Schlittenlacher) between the projects listed for it. Agreement between a project and the code it was ported or adapted from only checks the port.

• ported or adapted from another project
• the author's own code
• legacy legacy project

Scroll the picture sideways.

Implementations

Projects in bold and other established projects come first, then newly released, developing and legacy ones. Validation is what each project says about its own testing. The details are below, and the kinds of evidence are explained on the About page. The project map above shows which project was ported or adapted from which.

See also: Moore–Glasberg loudness of stationary sounds (ISO 532-2, ANSI S3.4).

How they were validated

As reported by each project; LISQM has not run the code. Two implementations that agree can still share the same error. Kinds of evidence.

iso532-3 (tv2018.m) · ISO 532-3:2023

not stated

• The folder "code for tables" has scripts that compute Table 5 (sone to phon) and the values of Annexes B and C (AnnexC.m is headed "Signals numbered as in ISO 532-2 Annex B"); the scripts do not name ISO 532-3.

AMT · Moore et al. (2016)

not stated

• There is no exp_moore2016 experiment script on the develop branch.

AMT · Glasberg & Moore (2002)

standard or paper data

• Has the experiment script exp_glasberg2002; AMT's experiment scripts reproduce the results of the publication they are named after.

AARAE · Glasberg & Moore (2002)

not stated

Ported or adapted from PsySound3 (the author's own code).

phonometry new · ISO 532-3:2023

standard or paper data

• A documented deviation: the Annex B WAV-file cases are not reproduced, and Annex C values are used instead.

torch_amt new · Moore et al. (2016)

torch_amt new · Glasberg & Moore (2002)

compared with AMT (its source)

Ported or adapted from AMT. It was compared with the code it was ported from, which checks the port only.

• Agreement with AMT is claimed in the documentation, without stored reference data.

LoudnessModel developing · Moore et al. (2018)

own tests only

Ported or adapted from iso532-3 (tv2018.m).

• Tests run in GitHub Actions compare overall loudness, the first five short-term and long-term values and selected intermediate stages for one input (1 kHz tone, 100 ms, 32 kHz) with hard-coded expected values.
• The README calls this "a known accurate output", but neither it nor the tests say where the values come from.

deeuu/loudness legacy · Glasberg & Moore (2002)

own tests only

No further details are recorded here.

HAW loudness tool legacy · Glasberg & Moore (2002)

reference code: LOUDAES and TVL (Glasberg & Moore's programs)

• The thesis compares the total loudness of pure tones (30–100 dB) with LOUDAES, Glasberg & Moore's program for stationary sounds: mostly within 5 %, with larger deviations at low levels near 50 Hz and above 5 kHz (−9.3 % for 5 kHz at 30 dB).
• For two equal-level tones at 1 and 2 kHz, the deviation from LOUDAES ranges from −4.4 % to +3.8 % with the Hann-window FFT and reaches +6.7 % with the Fourier-time transform.
• For the stationary part of signals the thesis reports good agreement with TVL, Glasberg & Moore's program for time-varying sounds, without figures.

Not stated by NumpyLibforPsychoAcoustic, PsySound3, and Swift & Gee codes.

Before you compare numbers

Implementations of the same metric can give different numbers without either being wrong. Check these choices first.

• Sample rate. The Cambridge programs and ISO 532-3 work at 32 kHz. Most implementations resample to it; phonometry runs at the native rate (a documented deviation), and the Swift & Gee code and the HAW tool expect 32 kHz input.
• Which output. The models give instantaneous, short-term and long-term loudness; compare like with like.

What the projects state:

phonometry new

• Documented deviation: it runs at the native sample rate instead of resampling to 32 kHz.

HAW loudness tool legacy

• Input must be sampled at 32 kHz; the reference level is entered as the sound pressure level of a full-scale sine.

Swift & Gee codes legacy

• The code works at 32 kHz; other rates are resampled with the Signal Processing Toolbox's resample.

References

1. Glasberg, B. R., & Moore, B. C. J. (2002). A model of loudness applicable to time-varying sounds. Journal of the Audio Engineering Society, 50(5), 331–342. https://aes2.org/publications/elibrary-page/?id=11081
2. Moore, B. C. J., Glasberg, B. R., Varathanathan, A., & Schlittenlacher, J. (2016). A loudness model for time-varying sounds incorporating binaural inhibition. Trends in Hearing, 20, 2331216516682698. https://doi.org/10.1177/2331216516682698
3. Moore, B. C. J., Jervis, M., Harries, L., & Schlittenlacher, J. (2018). Testing and refining a loudness model for time-varying sounds incorporating binaural inhibition. The Journal of the Acoustical Society of America, 143(3), 1504–1513. https://doi.org/10.1121/1.5027246
4. ISO 532-3:2023. Acoustics — Methods for calculating loudness — Part 3: Moore-Glasberg-Schlittenlacher method. https://www.iso.org/standard/69856.html
5. ISO/CD 532-3. Acoustics — Methods for calculating loudness — Part 3: Moore-Glasberg-Schlittenlacher method (2nd edition). https://www.iso.org/standard/90816.html