Why Loudness And Volume Are Different Measurements

Two recordings with identical peak levels can differ dramatically in how loud they seem. Perception depends on more than the height of the waveform.

Hearing is not equally sensitive across frequencies

The ear responds most strongly in the range where speech sits and less strongly at the extremes. Equal energy at different frequencies does not produce equal loudness.

A recording with substantial low-frequency content can measure high while sounding subdued. One weighted toward the midrange sounds louder at the same measured level.

Loudness measurement therefore applies a weighting curve that approximates the ear's response before averaging.

Duration matters as much as level

A brief peak contributes little to perceived loudness, while a sustained passage at a lower level dominates. Hearing integrates energy over time.

Loudness measures average over a program and discount very quiet passages so that silence between scenes does not drag the figure down.

The result is a single number describing how loud a whole program seems, which is what a viewer adjusting the remote actually cares about.

Normalization to peak does not equalize perception

Raising every file until its loudest instant touches the ceiling aligns peaks and nothing else. A dynamic recording still sounds quieter than a compressed one.

This is the mechanism behind decades of escalating loudness in music production, where heavy compression raised average level while peaks stayed fixed.

Loudness-based normalization removes the incentive, because a heavily compressed track is simply turned down to match.

Delivery platforms target a loudness level

Broadcasters and streaming services specify a target loudness for delivered programs, along with a limit on true peak to prevent clipping after encoding.

Material above the target is turned down; material below may be turned up or left alone depending on the platform. The details vary and change over time.

Mixing far louder than the target achieves nothing, since the excess is removed. It only sacrifices dynamic range that could have been used.

Dynamic range remains a creative decision

Loudness normalization equalizes averages, so a mix with wide dynamics keeps its quiet passages quiet and its loud ones loud.

Whether that suits the audience depends on listening conditions. Wide dynamics reward attentive listening and frustrate viewers in noisy rooms or on small speakers.

Some delivery chains offer alternative mixes with reduced range for exactly this reason. The choice is about the listening environment rather than about measurement.