Skip to content

Acoustic and Loudness Meters

Jump to bottom
youtube-at-vach edited this page Jul 23, 2026 · 1 revision

Acoustic and Loudness Meters

Relevant source files

The following files were used as context for generating this wiki page:

This page details the implementation of MeasureLab's acoustic, loudness, and precision timing instruments. These modules provide real-time monitoring of Sound Pressure Level (SPL), EBU R128/ITU-R BS.1770 loudness, noise floor characterization, and high-precision frequency/clock stability analysis.

1. Sound Level Meter (SPL)

The SoundLevelMeter provides real-time SPL monitoring with standardized frequency and time weightings. It supports A, C, and Z (Linear) frequency weightings and FAST (125ms), SLOW (1s), and IMPULSE time constants.

Technical Implementation

  • Filtering Pipeline: Input samples are processed through a cascaded IIR filter topology. A high-pass filter (20Hz) is always applied to remove DC and sub-sonics src/gui/widgets/sound_level_meter.py:181-189. Frequency weighting (A/C) is applied using SOS (Second-Order Section) filters designed via AudioCalc src/gui/widgets/sound_level_meter.py:196-202.
  • Time Weighting: Uses an exponential moving average (EMA) on the squared pressure values. The impulse weighting is unique, featuring a fast rise time (35ms) and a slow decay (1.5s) src/gui/widgets/sound_level_meter.py:79-82.
  • LN Statistics: Maintains a circular buffer of 0.1s samples to calculate percentile-based statistics (L5, L10, L50, L90, L95) src/gui/widgets/sound_level_meter.py:53-59.

Sound Level Meter Data Flow

graph TD
    A["AudioEngine (indata)"] --> B["Bandwidth Filter (20Hz-20kHz)"]
    B --> C["Frequency Weighting (A/C/Z)"]
    C --> D["Square & Integrator"]
    D --> E["Time Weighting (FAST/SLOW/IMPULSE)"]
    E --> F["dB Conversion & Calibration Offset"]
    F --> G["Lp (Instantaneous)"]
    F --> H["Leq (Equivalent Continuous)"]
    F --> I["LN Statistics (Percentiles)"]
Loading

Sources: src/gui/widgets/sound_level_meter.py:32-105, src/core/analysis.py:15-26

2. LUFS Meter (BS.1770)

The LufsMeter implements the ITU-R BS.1770-4 standard for loudness measurement, providing Momentary (400ms), Short-term (3s), and Integrated loudness metrics.

Key Components

  • K-Weighting: A two-stage filter consisting of a "Pre-filter" (high-frequency shelf) and a "RLB filter" (high-pass) src/gui/widgets/lufs_meter.py:105-120.
  • Gated Integration: The Integrated Loudness uses a dual-gate approach. An absolute gate at -70 LUFS is followed by a relative gate 10 dB below the ungated mean src/gui/widgets/lufs_meter.py:75-76.
  • Loudness Range (LRA): Calculated using the 10th and 95th percentiles of gated 3-second blocks src/gui/widgets/lufs_meter.py:171-186.

Sources: src/gui/widgets/lufs_meter.py:29-93, src/gui/widgets/lufs_meter.py:140-190

3. Noise Profiler

The NoiseProfiler characterizes the noise floor of a system, calculating Power Spectral Density (PSD) and identifying noise types (e.g., 1/f pink noise vs. white noise).

Implementation Details

  • PSD Calculation: Computes the magnitude spectrum in $V/\sqrt{Hz}$ using a Hanning window and appropriate scaling factors to ensure power conservation src/gui/widgets/noise_profiler.py:113-127.
  • Averaging Modes:
    • Exponential: Fixed alpha smoothing for real-time responsiveness src/gui/widgets/noise_profiler.py:84-88.
    • Cumulative: Linear average of $N$ spectra for high-precision stationary noise measurement src/gui/widgets/noise_profiler.py:68-77.
  • LNA Compensation: Allows users to input external LNA gain (dB) to refer noise measurements back to the DUT input src/gui/widgets/noise_profiler.py:151-154.

Sources: src/gui/widgets/noise_profiler.py:32-91, src/gui/widgets/noise_profiler.py:111-166

4. Frequency Counter and Allan Deviation

The FrequencyCounter provides sub-Hz precision frequency tracking and stability analysis.

Algorithms

  • Frequency Estimation: Uses calculate_frequency_metrics from core.frequency_analysis to perform high-precision estimation (likely via zero-crossing interpolation or phase slope) src/gui/widgets/frequency_counter.py:114-115.
  • Allan Deviation: Measures frequency stability over time. A QRunnable worker (AllanWorker) computes the deviation across multiple observation intervals ($\tau$) src/gui/widgets/frequency_counter.py:43-91.
  • Warmup Handling: Discards the first few valid points to prevent transients (e.g., DC blocking filter settling) from polluting long-term statistics src/gui/widgets/frequency_counter.py:139-140.

Sources: src/gui/widgets/frequency_counter.py:122-162, src/core/frequency_analysis.py:27-28

5. 1PPS Monitor

The OnePPSMonitor is an experimental tool designed to measure the drift of an audio interface's sample clock relative to a 1PPS (Pulse Per Second) reference (e.g., from a GPSDO).

Online Regression Engine

To estimate the sample rate with extreme precision, the module implements an online least-squares regression: $$SampleIndex = m \cdot PulseCount + c$$ Where $m$ represents the actual sample rate.

  • Trigger Logic: Uses a Schmitt trigger (Threshold + Hysteresis) to detect pulse edges in the audio stream src/gui/widgets/one_pps_monitor.py:44-45.
  • Robust Statistics: Employs Median Absolute Deviation (MAD) for outlier rejection to ignore pulses corrupted by electrical noise src/gui/widgets/one_pps_monitor.py:50-52.
  • Latency Compensation: Can automatically subtract the hardware input latency reported by the AudioEngine src/gui/widgets/one_pps_monitor.py:51-53.

Precision Monitoring Architecture

graph TD
    subgraph "Audio Callback"
        A["Audio In"] --> B["Trigger Detection (Schmitt)"]
        B --> C["Sample Timestamping"]
    end
    
    subgraph "Processing Thread"
        C --> D["Outlier Rejection (MAD)"]
        D --> E["Online Least-Squares Regression"]
        E --> F["Effective Sample Rate (m)"]
        E --> G["PPM Drift Calculation"]
    end
    
    subgraph "Data Storage"
        F --> H["instant_ppm_buffer"]
        G --> I["cumulative_ppm_buffer"]
    end
Loading

Sources: src/gui/widgets/one_pps_monitor.py:30-101, src/gui/widgets/one_pps_monitor.py:113-174, docs/widgets/one_pps_monitor.md:7-22

Summary of Key Classes

Class File Responsibility
SoundLevelMeter src/gui/widgets/sound_level_meter.py:32 SPL, A/C weighting, FAST/SLOW integration.
LufsMeter src/gui/widgets/lufs_meter.py:29 BS.1770-4 Loudness (M, S, I), LRA.
NoiseProfiler src/gui/widgets/noise_profiler.py:32 PSD, noise floor characterization, LNA gain.
FrequencyCounter src/gui/widgets/frequency_counter.py:122 Precision frequency tracking, Allan Deviation.
OnePPSMonitor src/gui/widgets/one_pps_monitor.py:30 Clock drift measurement via 1PPS pulse timing.

Sources: src/gui/widgets/sound_level_meter.py, src/gui/widgets/lufs_meter.py, src/gui/widgets/noise_profiler.py, src/gui/widgets/frequency_counter.py, src/gui/widgets/one_pps_monitor.py

MeasureLab Wiki

Home

Clone this wiki locally