The listener is part of the result
Loudness is a perceptual attribute. Signal amplitude contributes to it, but ears do not respond equally to every frequency, and that response changes with level. A low sine wave and a mid-frequency sine wave with the same measured pressure need not sound equally loud. Your headphones or speakers add another frequency-dependent influence before the sound reaches you.
Evidence: 1
A comparison has several possible controls
Matching the highest peak prevents one signal from having a larger maximum sample value. Matching RMS compares average squared amplitude over a chosen time window. Matching perceived loudness requires a listener judgement or a perceptual model with stated assumptions. These controls answer different questions. No single number guarantees that two differently shaped or differently timed sounds will feel equally strong.
Evidence: 2
Do not mistake stronger for richer
Suppose a harmonic mixture is compared with a sine wave. If the added components also make the signal much larger, the louder presentation may attract attention before you notice the spectral change. Keep playback modest and use the laboratory’s stated level controls. Listen for a specific attribute, such as the presence of a buzzy edge, instead of asking only which version sounds better.
Evidence: 3
Include silence as an option
The learning task remains meaningful without hearing every example. Compare the listed frequencies, relative component amplitudes and timing values. These show what the model changed; they cannot predict your exact experience. If a tone is faint, do not chase it by repeatedly increasing volume. Stop or choose a comfortable middle register and use the visual explanation for the remaining comparison. This is an educational site, not a hearing assessment.
Evidence: 1
Level matching removes one confound; it does not make perception identical.
Make the connection
Build a sound
Compare a sine-like and richer mixture at the same master volume, using the displayed level convention.
Open the experiment ↗Useful words
RMS
Root mean square: square the samples, average them, then take the square root. It describes signal magnitude but is not a complete loudness model.
See in glossary ↗Sources & further reading
These sources support the explanations and model assumptions. Recording credits are listed in the listening panels and the source register.
- Equal loudness contours and audiometry ↗Joe Wolfe / UNSW Music Acoustics · Undated
- Quantization and decibels ↗Chris Donahue / Carnegie Mellon University · 2026
- What is a sound spectrum? ↗Joe Wolfe / UNSW Music Acoustics · Undated