Target Curves Explained

Correction software measures your in-room response and generates filters to move it toward a reference shape. That shape is the target curve. Everything the correction does is relative to it, so the target embodies the answer to a question that sounds trivial and is not: what should a good system measure like in a room? The instinctive answer, flat from bottom to top, is wrong, and understanding why prevents the most common self-inflicted correction failure.

A loudspeaker's own design goal is a flat direct sound, measured anechoically on axis, with smooth behaviour off axis. Put that speaker in a room and the microphone no longer sees only direct sound. It sees direct sound plus everything the room returns, and the room returns more energy at low frequencies, where absorption is scarce and the speaker radiates in all directions, than at high frequencies, where directivity narrows and every soft surface absorbs. A steady-state measurement of a genuinely neutral system therefore tilts downward from bass to treble, typically in the region of half a decibel to a decibel per octave depending on the room and the speaker's directivity.

The ear develops in rooms and expects this. It hears the direct sound as the timbre reference and folds the room's contribution in as a sense of space and warmth. So when correction forces the steady-state measurement flat, it is not making the system neutral, it is tilting a neutral system upward: cutting the bass the room legitimately contributes and boosting treble until the direct sound is bright. The result measures perfectly and sounds thin, and it is the most common outcome of naive correction. The frequency response guide covers the related distinction between direct and steady-state response.

The research record supports the tilt. The Harman listening studies, among others, found listeners consistently prefer an in-room response that slopes gently downward, with a modest low-frequency shelf, and the pattern held across trained and untrained listeners. The debated territory is the size of the bass shelf and the exact slope, which is taste and program dependent, not whether the curve tilts.

The purpose of the system moves the sensible target. A mixing or mastering room is a measuring instrument: decisions made there must translate, so the target stays conservative, the natural tilt with little or no added bass shelf, because a bass-rich monitoring target reads as a bass-light mix everywhere else. A hi-fi or listening room carries no translation duty, and a couple of decibels of additional low shelf is a legitimate preference rather than an error. Same physics, different stakes.

The other constraint on the target is the hardware. A target is an instruction set, and asking for extension the speakers do not have converts the instruction into boost, spent as amplifier power and driver excursion at frequencies the cabinet cannot reproduce. Small ported monitors roll off steeply below their tuning, and correction boost below that point produces distortion and port noise, not bass. Set the target's low-frequency limit at what the system genuinely delivers, and if more extension is wanted, that is a subwoofer decision, not a curve decision. Correction software with restraint built in will refuse the boost anyway, which is a feature, not a limitation.

Start with the product's default tilted target, correct, verify, and listen to material you know deeply for several days before touching anything. If a change still feels warranted, adjust the low shelf in decibel steps and re-verify, and leave the top octaves alone, since steady-state treble at a microphone is a poor guide to perceived brightness and the direct sound there is already what the speaker's designer intended. Distrust any urge to make the measured line straighter for its own sake. The measurement serves the listening, not the reverse.