SPL Measurement per IEC 61672-1: Complete Guide

Understanding IEC 61672-1 Sound Level Meters

IEC 61672-1 is the international standard that specifies the performance requirements for sound level meters. It defines two accuracy classes:

• Class 1 (Precision): ±1.1 dB tolerance at reference frequency. Required for legal noise monitoring and workplace compliance assessments.
• Class 2 (General Purpose): ±1.4 dB tolerance. Suitable for environmental surveys, sound system tuning, and general noise assessment.

SonaVyx's browser-based SPL meter implements Class 2 accuracy using calibrated A, C, and Z frequency weighting filters implemented as IIR biquad cascades via bilinear transform from the analog prototype curves defined in the standard.

Frequency Weighting: A, C, and Z

The standard defines three frequency weighting networks:

• A-weighting (dBA): Approximates human hearing sensitivity. Most commonly used for noise assessment, hearing conservation, and environmental monitoring. De-emphasizes low frequencies below 500 Hz and high frequencies above 6 kHz.
• C-weighting (dBC): Nearly flat response. Used for peak measurements (LCpeak) and assessing low-frequency noise content. Required by EU Directive 2003/10/EC for peak exposure assessment.
• Z-weighting (dBZ): Flat response (zero weighting) from 10 Hz to 20 kHz. Used when unweighted measurements are needed for engineering analysis.

Time Weighting

IEC 61672-1 §5.5 specifies three time weighting characteristics:

• Fast (F): 125 ms time constant. Good for capturing moderately varying sounds.
• Slow (S): 1000 ms time constant. Smooths out rapid fluctuations for stable readings.
• Impulse (I): 35 ms attack, 1500 ms decay. Captures impulsive sounds like hammering or gunshots.

Measurement Procedure

1. Calibrate: Use a known 94 dB calibrator at 1 kHz. In SonaVyx, go to Settings → Calibration and follow the on-screen procedure.
2. Position: Place the microphone at the measurement position, 1.2-1.5m height, away from reflecting surfaces (per IEC 61672-1 §6.2).
3. Select weighting: Choose A-weighting for general noise assessment, C-weighting for peak measurements.
4. Measure: Record for at least 30 seconds for steady-state noise, longer for fluctuating environments.
5. Document: Note Leq, Lmax, Lpeak, and statistical percentiles (L10, L50, L90).

SonaVyx Implementation Details

SonaVyx implements the IEC 61672-1 frequency weighting filters using cascaded IIR biquad sections. The filter coefficients are computed via bilinear transform from the analog prototype defined in Annex E of the standard. Time weighting is implemented as a first-order exponential averaging filter in the Rust WASM DSP engine, processing audio at 48 kHz sample rate with 32-bit float precision.

Key metrics computed in real-time:

• Leq: Equivalent continuous sound level — the steady level with the same energy as the fluctuating measured level.
• Lmax/Lmin: Maximum and minimum time-weighted levels during the measurement period.
• Lpeak: True peak sound pressure level (unweighted, or C-weighted as LCpeak).
• L10/L50/L90: Statistical percentiles from a 0.1 dB resolution histogram.