Skip to content

Imaging: How a System Places Instruments in Space

Imaging is how precisely and how stably a system places individual instruments within the space it reproduces. It is separate from soundstage, which is the size of that space: a system can present a huge stage with vague placement, or a modest one where every instrument sits in a fixed, pointable position.

Work it out for your gear

Online Tone Generator Pure sine waves, white, pink, brown, blue, and violet noise. Frequency sweeps, channel test, and polarity check. No download. Runs entirely in your browser.

The two cues everything rests on

Your brain locates a sound with two differences between what your left ear hears and what your right ear hears.

The first is timing. A sound from your left reaches the left ear slightly before the right, by up to about 700 microseconds at the extreme. This is the dominant cue below roughly 1 kHz, where the wavelength is longer than the width of your head and the phase difference is unambiguous.

The second is level. Above about 1.5 kHz your head casts an acoustic shadow, so the far ear receives less energy than the near one. Level difference takes over as the dominant cue in that region.

Between those two lies a band around 1 to 2 kHz where neither cue works especially well, which is one reason localisation is least precise there. On top of both sits the filtering your outer ear performs, which varies with the direction a sound arrives from and is what lets you tell front from back and judge elevation at all.

Everything a system does to imaging, it does by preserving or corrupting those differences.

What corrupts them

Polarity. If one channel is wired out of phase, the difference cues become contradictory: your brain gets a signal that cannot correspond to any real source position. Centre images collapse, bass partly cancels, and the presentation sounds oddly wide and hollow with a hole in the middle. This is the single most common serious imaging fault and it is entirely fixable, which is why it is the first thing to rule out.

Channel imbalance. A level difference between channels shifts every image toward the louder side. Small amounts pull the centre off. This is worth checking before blaming a component, because a volume control with poor channel tracking at low settings is common, and so is a headphone with one pad noticeably more compressed than the other.

Crosstalk. Signal leaking from one channel into the other reduces the difference between the two, which narrows and blurs placement. In a well built modern component it is far below the level where it matters. In a turntable cartridge, a long unbalanced run in a noisy environment, or a poorly implemented shared ground return, it is not.

Resonance and ringing. Placement depends on fine timing detail, and a driver or enclosure that continues to ring after the signal stops smears exactly that detail. This is why transducers described as fast or clean are so often described as well separated too: it is one property being named twice.

The room, for loudspeakers. A strong early reflection arriving within a few milliseconds of the direct sound competes with it for the same localisation judgement. Treating the first reflection points and getting the two speakers symmetric relative to the side walls typically produces a larger improvement in imaging than any equipment change at the same cost.

Why headphones image differently

Headphones deliver each channel to one ear with none of the outer-ear filtering that a real source in a room would impose, and with no crosstalk at all, since the left driver is not audible to the right ear. The result is unnaturally high channel separation with the wrong spatial cues attached to it.

That combination is why headphone images tend to sit inside the head and to lie along a line between the ears rather than out in front. It also explains a genuine strength: because there is no room and no crosstalk, headphones resolve small differences between instruments better than most speakers do, even while placing them less believably.

Open-back designs, angled drivers and crossfeed all give back some of the missing cues and move the image outward to a degree. Binaural recordings, which capture the ear filtering at the source, come closest of all, and are worth hearing once for the demonstration alone.

Testing your own setup

The tone generator below has the two tests that matter, and they take about five minutes.

Run the channel test first. Each channel should be clearly and only on its own side, and the two should sound equally loud. Then run the polarity check: with correct wiring a centred low tone sounds solid and located between the speakers or between your ears, and with one channel inverted it sounds diffuse and hard to place.

After that, use music you know well. A well recorded solo voice should sit in one stable spot and not wander as it changes pitch. A drum kit should keep its parts in fixed places rather than swelling toward the centre when the mix gets busy. If the centre image drifts when you move your head slightly, you are hearing the room, not the speakers.

Reading the word in a review

Imaging claims are only useful when they name what the comparison was and what was playing. Precision also has a ceiling set by the recording: a hard-panned studio mix contains little placement information, and no component can invent it. When two reviewers disagree about a headphone's imaging, the most common explanation is that one of them was describing soundstage.

Reviewed gear that shows this

14 reviews in the catalogue, highest-scoring 6 shown. Every score is anchored to the same published reference list.

Keep reading