“Fletcher–Munson” and “Zwicker loudness” are often used in the same conversation because both concern how humans hear sound. They answer different questions. One describes the sound-pressure levels needed for pure tones to feel equally loud; the other estimates the loudness of an entire stationary spectrum.
What are Fletcher–Munson curves?
Fletcher–Munson curves are a historical name for equal-loudness contours. Each contour shows combinations of frequency and sound-pressure level judged equally loud to a reference tone at 1 kHz. They illustrate a familiar fact: bass and very high frequencies usually need more physical SPL than midrange sound to be perceived at the same loudness.
Modern equal-loudness standards refine those early data, but “Fletcher–Munson curves” remains a widely used name for the general idea. The curves are especially useful for teaching frequency-dependent sensitivity and for understanding why a flat SPL response is not necessarily perceived as flat.
What does stationary Zwicker loudness add?
Zwicker loudness starts with a multi-band spectrum rather than one isolated tone. It considers auditory critical bands, level-dependent behaviour, threshold-related effects, and masking between nearby auditory regions. Its output can therefore describe both total loudness in sones and the distribution of loudness across the Bark scale.
That does not mean it replaces an equal-loudness contour. Instead, it tackles a different practical problem: given a declared spectrum, how loud is the whole sound likely to be perceived?
| Concept | Primary question | Typical output |
|---|---|---|
| Equal-loudness contour | What SPL is needed at each frequency to match a 1 kHz reference in loudness? | A family of frequency-versus-SPL curves, labelled in phons. |
| A-weighting | How can a sound-level meter apply a practical fixed frequency weighting? | A weighted level in dBA. |
| Stationary Zwicker loudness | How loud is this entire calibrated spectrum, and where is its loudness distributed? | Total sones, loudness level in phons, and specific loudness in sone/Bark. |
Why A-weighting is not Zwicker loudness
A-weighting is a fixed curve applied to a sound-pressure measurement. It is fast and useful for many practical noise measurements, but it does not reproduce the full level- and spectrum-dependent behaviour of a loudness model. It cannot show a specific-loudness pattern, and two spectra that have the same dBA value can still differ in perceived loudness.
How this applies to a frequency-response screenshot
A response screenshot often uses a relative scale such as “0 dB at 1 kHz.” That chart can be excellent for comparing spectral balance, but it lacks the absolute acoustic level needed for either a meaningful sone value or a loudness-level result in phons. Select Uncalibrated response in the calculator to digitize it without generating a misleading loudness result.
If the graph is a transfer function and you know the actual SPL at 1 kHz, select Transfer / relative response and enter that reference. The calculator subtracts the trace’s actual 1 kHz value, then adds the declared SPL anchor. That reconstruction is labelled as an estimate because it is not the same as measuring one-third-octave band energies directly.
When should you use each idea?
- Use equal-loudness contours to understand frequency sensitivity and how a pure tone may need different SPL at different frequencies.
- Use A-weighting when a measurement requirement asks for dBA or when a fast broad comparison is appropriate.
- Use stationary Zwicker loudness when you have a calibrated spectrum and need a human-perception estimate in sones, phons, and Bark bands.
To go deeper into the calculation itself, read Zwicker loudness: from sound pressure level to sones. To understand its detailed output, see specific loudness in sone/Bark.