For most noise measurements, A-weighting is the right choice. It’s the standard required by workplace safety regulations, environmental noise assessments, and community noise ordinances. But if you’re measuring loud impulsive sounds, dealing with heavy low-frequency noise, or need a completely unfiltered reading, a different weighting may serve you better. The answer depends on what you’re measuring and why.
What Frequency Weightings Actually Do
Your ear doesn’t treat all frequencies equally. Low rumbles need to be much louder than mid-range sounds before you perceive them at the same volume. Frequency weightings are filters built into sound level meters that adjust raw measurements to account for this uneven sensitivity. Without a weighting, you’d get a technically accurate number that doesn’t reflect how loud something actually sounds to a person.
Three weightings matter today: A, C, and Z. Two others, B and D, were designed for moderate sound levels and aircraft noise respectively, but neither appears in modern standards and you’re unlikely to encounter them outside of legacy equipment.
A-Weighting: The Default for Most Situations
A-weighting reduces low frequencies and slightly boosts the 1 to 4 kHz range, mimicking how the human ear responds to sound at conversational to moderate levels. It was developed from the inverse of the 30-phon equal loudness curve, which represents the threshold at which sounds across different frequencies are perceived as equally loud at a relatively quiet level. In practice, it treats bass rumble as less significant than mid- and high-frequency noise, because that’s genuinely how your hearing works at typical volumes.
This is the weighting mandated by virtually every occupational and environmental noise regulation. OSHA requires all workplace noise exposure measurements to use the A scale, with permissible exposure set at 90 dBA over an 8-hour day and a hearing conservation trigger at 85 dBA. NIOSH similarly specifies A-weighting with slow response for both sound level meters and dosimeters. Community noise metrics like the Day-Night Average Sound Level (Ldn) and the WHO’s environmental noise guidelines all use A-weighted measurements.
If you’re measuring workplace noise, neighborhood noise complaints, HVAC systems, traffic, construction equipment, or general environmental sound, A-weighting is what you should use. It’s also the weighting that will match published noise limits in local ordinances and building codes.
C-Weighting: When Bass Matters
C-weighting follows the 100-phon equal loudness curve, which represents how your ear perceives frequency at very loud levels. At high volumes, your hearing flattens out considerably: bass doesn’t need nearly as much of a boost relative to mid-range to sound equally loud. So C-weighting applies a much flatter filter than A-weighting, keeping far more of the low-frequency energy in the measurement.
You should use C-weighting in two main scenarios. First, when measuring peak sound pressure levels from impulsive or explosive sounds, like gunfire, hammering, or industrial stamping presses. These short, intense bursts of noise can cause immediate hearing damage, and A-weighting would undercount the low-frequency energy that contributes to that risk. Second, C-weighting is standard for machine noise and jet noise measurements where low-frequency content is a dominant part of the sound profile.
A useful diagnostic trick: if you measure the same sound with both A and C weightings and the C reading is significantly higher (more than about 10 dB), the noise has a strong low-frequency component that A-weighting is filtering out. This gap tells you low-frequency energy may be a problem even if the A-weighted number looks acceptable, which is common with large industrial machinery, ventilation systems, and music venues with heavy bass reinforcement.
Z-Weighting: The Unfiltered Reading
Z-weighting, defined in the IEC 61672 standard, applies no frequency adjustment at all. It provides a flat response from 8 Hz to 20 kHz within plus or minus 1.5 dB, capturing exactly what the microphone picks up. The “Z” stands for zero, as in zero human-ear correction.
This is the weighting for engineering and diagnostic work where you need the actual acoustic energy present, not a human-perception-adjusted version. If you’re performing octave band analysis, troubleshooting equipment vibration, feeding data into acoustic modeling software, or comparing raw sound power output between machines, Z-weighting gives you the uncolored data to work with. You can always apply A or C corrections mathematically afterward, but you can’t recover filtered-out frequencies from an A-weighted recording.
Quick Selection by Application
- Workplace noise compliance (OSHA, NIOSH): A-weighting, slow response
- Environmental and community noise: A-weighting
- Building acoustics and HVAC: A-weighting for general assessment, C-weighting if low-frequency rumble is suspected
- Peak and impulse noise (gunfire, impacts): C-weighting
- Loud machinery with dominant bass content: C-weighting, or both A and C for comparison
- Engineering analysis and raw data collection: Z-weighting
- Acoustic research and octave band work: Z-weighting
When You’re Still Unsure
If no regulation or standard is telling you which weighting to use, start with A-weighting. It’s the most universally accepted, the most comparable to published reference levels, and the best approximation of perceived loudness for the broadest range of everyday noise levels. You can always take a parallel C-weighted measurement to check for hidden low-frequency energy.
If you’re taking measurements that need to hold up to legal or regulatory scrutiny, confirm the specific weighting required by your local jurisdiction or the standard you’re working under. Nearly all will specify A-weighting, but peak exposure limits sometimes reference C-weighted values separately. Having both numbers in your records covers most situations.