How Do You Know a Loudspeaker Is Good?

A loudspeaker can have excellent specifications, beautiful construction, and impressive measurements. It can also sound spectacular during a five-minute demonstration. None of those things, by itself, tells us whether it is a good loudspeaker.

The final test is whether the loudspeaker reproduces music convincingly, consistently, and without imposing a persistent character of its own. Determining that requires listening, but useful listening is more disciplined than simply playing a favorite recording and deciding whether the sound is pleasant.

Listening and Measurement Are Partners

Listening is sometimes described as the subjective part of loudspeaker design, while measurement is considered the scientific part. That distinction is misleading. Measurements tell us what a loudspeaker is doing. Listening tells us what the audible consequences are. A measurement may reveal a resonance, irregular dispersion, or rising distortion, while listening tells us whether that behavior actually interferes with the music.

The opposite is equally important. Listening may reveal that something sounds wrong without explaining why. What initially sounds like too much treble may actually result from diaphragm breakup, a narrow resonance, poor crossover integration, or excessive reflected energy from the room. The ear reports the experience. Measurement helps explain it. When the two disagree, the proper response is not to choose sides but to investigate.

Firstly, Control the Conditions

Comparisons become unreliable if several things change at once. Loudspeaker position, listening distance, toe-in, amplifier, source equipment, and playback level can all alter what we hear.

Playback level is particularly deceptive. A slightly louder loudspeaker will often seem to have better bass, greater clarity, and stronger dynamics even when the listener does not consciously notice the difference in loudness. Meaningful comparisons therefore require careful level matching. The same principle applies during development. Change one meaningful variable, return to the same program material and level, and listen again. Otherwise, a loudspeaker may improve without revealing why it improved, which is not very useful engineering.

Listen to More Than One Kind of Music

No single recording can evaluate an entire loudspeaker. Different sounds reveal different weaknesses.

Human voice is exceptionally useful because we spend our lives hearing voices. Small colorations can become obvious, and a voice that changes character, size, or position as its pitch changes may reveal poor crossover integration.

Piano music tests an enormous portion of the system at once. It combines a wide frequency range, sharp attacks, complex decay, dynamics, and closely spaced pitches. Strings can reveal resonances and roughness. Brass exposes hardness and compression. Percussion tests transient response and recovery. Bass instruments reveal whether low frequencies retain actual pitch instead of becoming generalized thumps and rumbles. It is also useful to play ordinary recordings as well as superb ones. A loudspeaker that works beautifully only with a handful of demonstration-quality recordings is not especially useful. At the same time, a loudspeaker that makes every recording sound equally pleasant may simply be concealing information. A good system should reveal differences without making those differences punitive.

Listen for What Remains the Same

One of the most revealing questions is whether unrelated recordings begin acquiring the same character. Do all voices sound slightly chesty? Do all cymbals have the same metallic sheen? Does every recording possess unusually warm bass? Does every orchestra seem to occupy the same enormous space? If so, you may be hearing the loudspeaker rather than the recordings.

A neutral loudspeaker should allow recordings to remain different. A dry recording should sound dry. A reverberant hall should retain its reverberation. A close microphone should sound close. A distant ensemble should sound distant. Neutrality does not mean dullness. It means that the loudspeaker does not insist upon performing alongside every musician.

Listen to the Stereo Image

A centered monaural voice is one of the simplest and most revealing stereo tests. It should appear as a stable sound between the loudspeakers rather than as two voices coming from two boxes. Its position should not wander as pitch changes. Its apparent size should remain plausible, and the individual loudspeaker drivers should not become identifiable as separate sources.

Stereo recordings should produce a continuous panorama rather than isolated left, center, and right regions. Depth should change according to the recording. Some recordings should appear intimate and forward, while others may extend considerably behind the loudspeakers. If every recording produces spectacular depth and width, the system may be manufacturing an effect rather than reproducing one.

Move Around

A loudspeaker should not sound correct only when your head occupies one precisely marked location. Moving sideways, vertically, or changing listening distance will naturally alter the presentation somewhat. The important question is whether those changes occur gracefully. An abrupt tonal change caused by moving your head a few inches may reveal crossover lobing or irregular directivity.

Every multiway loudspeaker also has some distance at which its individual drivers become perceptually integrated. A near-field monitor should integrate at a short distance. A large theater system can reasonably require more. What matters is that the required listening distance suits the intended application. A domestic loudspeaker that needs a ballroom before its drivers become one source has a practical problem.

Turn It Up, and Turn It Down

A loudspeaker should retain its basic character over a useful range of levels. At low levels, listen for fine detail and the ability to preserve subtle changes. At moderate levels, tonal balance and integration become important. At higher levels, listen for compression, hardness, congestion, cabinet noise, port noise, and loss of bass pitch. A loudspeaker approaching its limits often announces the fact before obvious distortion appears. The stereo image may flatten, brass becomes aggressive, bass loses definition, or dynamic peaks stop growing naturally even though the volume continues increasing. Good dynamics should sound easy. The listener should hear the music becoming more powerful, not the machinery working harder.

The Five-Minute Test Is Not Enough

Some loudspeakers are designed, intentionally or otherwise, to make a strong first impression. Elevated treble can initially sound detailed. Emphasized bass can sound powerful. Extremely sharp image outlines can seem remarkably precise.

An hour later, those same characteristics may become exhausting. Listening fatigue is therefore useful information. A loudspeaker that continually directs attention toward some aspect of its own sound is likely doing something worth investigating. On the other hand, a system that never becomes tiring because it suppresses transients or information is not necessarily accurate either. The goal is a loudspeaker that remains revealing without becoming irritating. Long listening sessions tell us something short demonstrations cannot: whether we continue wanting to play music.

Does the Music Still Communicate?

There is one final test that is difficult to put on a specification sheet. Technical accuracy and emotional involvement are not opposites. If resonances, distortion, timing errors, and tonal imbalances are reduced, the musical distinctions through which emotion is communicated should become easier to hear. Phrasing, tension, release, dynamics, rhythm, and interaction among musicians should become clearer. The loudspeaker should not manufacture excitement, warmth, drama, or intimacy. Those qualities should change with the performance. A restrained performance should remain restrained. A powerful one should become powerful. The loudspeaker should not tell you what to feel. It should allow the musicians to do that.

When Is It Good?

A good loudspeaker survives more than one impressive demonstration. It works with varied music, realistic levels, long listening sessions, different positions, and reasonable rooms. Bass retains pitch, instruments retain recognizable timbre and scale, stereo images remain stable, dynamics grow without obvious strain, and different recordings remain distinctly different. Most importantly, after a while, the listener stops evaluating the loudspeaker. The equipment ceases to be the subject, and attention returns to the music.