As introduced in Understanding the Ear’s Sensitivity and Loudness Perception, human hearing varies in sensitivity across different frequencies and sound levels, which significantly influences how loud a sound feels.
This is why sound meters use frequency-weighting curves like A, B, C, and Z to adjust the signal before showing you a result.
In this post, we explain:.
- what these curves are;
- why they matter;
- and how they help us measure sound in a way that reflects human perception
Enjoy the read!
The origin of acoustic weighting curves
In Understanding the Ear’s Sensitivity and Loudness Perception, you learn that humans don’t hear all sounds the same way — we are more sensitive to certain frequencies than others.
(If you want to better understand the concept of frequencies, read this article).
So, when we measure sound, we need to account for how we actually hear it, not just what the microphone picks up.
This is done by applying what’s called a weighting to the sound signal.
The equal-loudness curves (also Fletcher-Munson curves) introduced previously are the foundation for this process. Remember, they look like this.
However, these curves are too complex to use directly in sound measuring devices.
The original contours vary too much to build directly into the hardware. To simplify things, engineers flipped the curves upside down and smoothed them into easy-to-use electronic filters:
- A-weighting ≈ inverse of the 40-phon curve (normal ‘everyday’ sounds)
- B-weighting ≈ inverse of the 60-phon curve (medium-loud)
- C-weighting ≈ inverse of the 80-phon / 100-phon curve (very loud)
- Z-weighting = A flat, unmodified response – no boost, no reduction (think “Zero”).
(If you want to know where to find the exact values used in each third-octave and octave frequency band, scroll down to the bottom of the post)
Note
the scale is now using decibels (dB), which we explain in the next post
A closer look at the acoustic frequency weighting curves
A-weighting – Standard curve
A-weighting mainly reduces low frequencies and slightly boosts some mid-high frequencies.
It is designed to match how we perceive sounds at normal levels — usually around 45 to 85 dB.
Today, A-weighting is used almost everywhere:
- in building regulations
- product noise data
- and standards.
If you have ever seen dB(A) on a report or a sound level meter, this is the A-weighting being used.
B-weighting – Mid level curve
B-weighting was designed for medium-loud sounds. It cuts the low frequencies less than A-weighting but isn’t widely used anymore.
You might still find it in older research papers or measurement reports, especially from the 1970s and 1980s.
C-weighting – Mid level curve
C-weighting provides a nearly flat response across most of the hearing range, especially keeping the low frequencies that A-weighting removes.
Z-weighting – Raw truth
Z-weighting is completely flat from 10 Hz to 20 kHz, with no correction applied. It represents the raw, unfiltered signal that the microphone picks up.
Use dB(Z) when you want to see the true frequency content of the sound, without any adjustments.
How acoustic weighting is applied in practice
When you — or a sound level meter — analyse a sound, you can either look at the raw data or apply a specific weighting:
- dB(A)
- dB(B)
- or dB(C)
If you choose a weighting, the device processes each frequency band individually, applying corrections based on the gain of the selected curve.
It adds or subtracts gain values band-by-band, depending on how sensitive the human ear is to each frequency.
“The result is a single, easy-to-understand value that combines all frequency bands. We call this a ‘broadband‘ value, and it gives a better sense of how loud a sound feels to the human ear.”
The letter in brackets — A, B, or C — tells you which weighting curve was used.
Note
You can find the correction values for each weighting curve in the following standards:
for A and C weighting curves, IEC 61672 – Electroacoustics – Sound level meters
for the historic B weighting curve, IEC 60651 – Sound level meters