DeviceBenchFree online speaker test
Speaker test that finds a box wired the wrong way round
Pink noise goes to the left box on its own, then to the right on its own, and you say where it came from rather than where you expect it to be. Two further passes follow that a left-then-right check never reaches: an alternating burst with a trim you slide until the two match, which turns a vague sense of imbalance into a figure in decibels, and a pair of sample-identical bursts with one channel’s sign flipped, which is the only way to hear a speaker connected with its two terminals reversed. There is no microphone here, so the page supplies the stimulus and your ears supply every verdict.
- 100% free
- No signup
- Pink noise
- ±12 dB trim
- Polarity pass
1. Which box is which
Pink noise, 2.5 seconds a side. Sit where you normally sit and answer from where the sound appeared, not from where you expect it to be.
Left channel only
The right output is held at silence, not merely turned down.
Right channel only
Same burst, opposite side. Answer before you look at the boxes.
The reference pass. With a correctly placed pair this should appear to come from a point between the boxes rather than from either one.
2. Is one side quieter
The same burst alternates between the boxes about once a second. Move the trim until the two feel equally loud, then keep the number: that is the imbalance, in decibels.
3. Is one box wired backwards
Two passes of identical noise, three seconds each. The second one has the right channel flipped in sign. On a healthy pair the flipped pass loses its middle and seems to come from beside your ears instead of from between the boxes.
Every pass is generated in the tab and goes wherever your system is currently sending sound, so a laptop is testing its own two drivers a few inches apart. A phone with a single speaker will report every pass as coming from both sides, correctly — plug in headphones or run this on the machine the speakers are attached to.
How to test a pair of speakers
Three passes in this order: which box is which, how far apart they are, and whether one is inverted.
Sit where you actually listen, then answer from where the sound appeared
Stand up and the whole test changes: a pair aimed at a chair puts its image somewhere else entirely if you are leaning over the desk. Play the left burst first and answer before you glance at the boxes — the point of the question is to catch the case where the box on your left is carrying the right channel, and knowing which button you pressed contaminates that.
Match the two levels by ear and write the number down
The second pass alternates a burst between the boxes about once a second and gives you a trim in half-decibel steps. Slide it until neither side pulls your attention, then read the figure: that is how far apart the pair is. Do it before you touch any tone control, because a treble knob at a different position on each box will show up here as an imbalance that follows the frequency rather than the channel.
Compare the two polarity passes back to back
Pass A sends identical noise to both boxes; pass B sends the same noise with the right channel's sign flipped. Play them one after the other without moving your head. On a correctly wired pair, B loses its middle and seems to widen out past the boxes. If the two passes are indistinguishable, one speaker is already wired backwards and the flip in pass B simply put it right.
Technical specifications
| What the page can observe | Nothing. It generates the signal and you supply the answer, so every verdict below is your ears plus a known stimulus |
|---|---|
| Routing signal | Pink noise, equal energy per octave, 2.5 s per channel |
| Channel isolation | StereoPannerNode at ±1, so the resting side receives no signal at all rather than a quieter copy |
| Trim range | ±12 dB in 0.5 dB steps, applied by attenuating the louder box so nothing gets louder than the reference |
| Alternation rate | A 0.9 s burst every 1.15 s, which keeps the comparison inside auditory memory |
| Polarity pair | 3 s of sample-identical noise in both channels, band-limited to 120-1400 Hz, right channel inverted for the second pass |
| Why that band | Below 120 Hz room modes imitate the effect on their own; above 1.4 kHz the wavelength is shorter than the gap between your ears |
| Audible imbalance | About 2 dB before the image visibly leans; 1 dB is inside the tolerance a matched pair is built to |
Frequently asked questions
Both boxes play, but everything seems to come from one side. What causes that?
A level difference between the channels, and about two decibels is enough to be obvious — the pass with the trim slider puts a number on exactly this. The three usual sources are a balance control that was nudged and forgotten, a volume knob whose two halves no longer track each other, and a speaker cable with one strand of copper still making contact where the rest has broken. Work them in that order: the first two cost nothing to check.
What does out of phase actually sound like?
Thin, wide and impossible to place — the sound seems to come from beside your ears or from behind the boxes rather than from a point between them, and the bass drops noticeably. What you are hearing is cancellation: two identical waves arriving with opposite signs subtract from each other in the air, most completely at the position exactly between the boxes, which is where the stereo image is supposed to be built.
How do I fix a speaker that is wired backwards?
Swap the two wires at one speaker only — the end that goes into the box, not the amplifier. Passive speaker cable is marked so you can tell the conductors apart: one side has a ridge you can feel, or a stripe, or printing along it, and the convention is that the marked conductor goes to the red terminal at both ends. Reversing both ends changes nothing, which is why the fault survives so many attempts to fix it.
Why pink noise rather than a tone?
Because a single tone tells you as much about your room as about your speakers. A tone at one frequency builds standing waves between the walls, so it can be loud at your chair and nearly silent a step to the left, and a channel can seem dead when it is merely sitting in a null. Pink noise carries equal energy in every octave, so it averages the room out and localizes far better — which is why it is also what installers use to align a system.
One box is quieter only when I turn the volume down. Why?
That is channel tracking error in an analog volume control, and it is a wear pattern rather than a fault in the speaker. A dual potentiometer has two resistive tracks that are supposed to move together, and they match worst near the bottom of their travel, where a mismatch of three or four decibels is common even in new equipment. Run this test at your normal listening level and again very quietly: an imbalance that appears only in the second run is the knob.
The test says both boxes played the same burst. What makes a stereo pair go mono?
Most often a plug that does not match its socket. A three-conductor jack pushed into a four-conductor headset socket, or the reverse, shorts a channel to the sleeve and the system falls back to sending the same signal everywhere. After that, check the accessibility settings: Windows, macOS, iOS and Android all have a switch that combines both channels into mono for people with hearing in one ear, and it is easy to turn on by accident.
Does this work for a laptop's built-in speakers or a soundbar?
For a laptop, yes, and the answers are usually less interesting: the two drivers are a few inches apart, so the routing pass works but the polarity pass has almost no image to collapse. A soundbar is the harder case — many of them apply their own widening or virtual-surround processing, which deliberately mixes some of each channel into the other and can make a genuinely dead channel still produce sound. Turn every sound mode off before you trust a result from one.
About polarity, balance and what a stereo pair is supposed to do
Two speakers are not two independent sound sources — they are one instrument that builds images between them. A sound mixed equally into both channels does not appear to come from either box; it appears to come from a point in the middle, because both ears receive it at the same instant and at the same level and the brain resolves that into a single source straight ahead. Everything a stereo pair does well depends on that phantom center holding still, which is why the two faults this page hunts for matter far more than their size suggests: a couple of decibels of imbalance drags the whole image toward one box, and an inverted speaker removes the center outright while leaving both boxes measurably working.
Polarity is the one people miss, because nothing looks or measures wrong. A passive speaker is driven by a two-conductor cable and the amplifier pushes the cone out when the signal goes positive; reverse the two conductors at one box and that cone now pulls in while its partner pushes out. Both boxes still play, at the correct level, with the correct tone. What is lost is the summation in the air between them: the two waves subtract instead of adding, most severely at low frequencies where the wavelength is long enough that both boxes reach your chair in step. That is why the test here is band-limited rather than full-range — below about 120 Hz a room’s own reflections produce cancellation of their own, and above roughly 1.4 kHz the wavelength is shorter than the distance between your ears, so the effect stops being reliable at all. Cable manufacturers mark one conductor with a ridge, a stripe or printing for precisely this reason, and the convention is that the marked side goes to red at both ends.
The neighboring pages here take the questions this one deliberately leaves alone. If neither box produced anything at all, the fault is upstream of the wiring and the sound test settles that in one press. How low the pair actually reaches before it starts distorting instead of playing is the bass test, and a steady tone you can hold at any frequency while you move around the room comes from the tone generator — walking a 60 Hz tone around a small room is the fastest way to see how much of what you blame on speakers is standing waves. Headphones need a different set of checks entirely, since there is no air path between the two drivers for anything to cancel in; those are on the headphone test. And when there are more than two boxes involved, each one has a numbered output behind it, which the surround sound test addresses one at a time.
Where the noise comes from
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type, paste or open is uploaded, logged or kept, which is also why the tools carry on working after you disconnect from the network.
The pink noise and the polarity pair are both synthesized sample by sample in this tab — there is no audio file to download and nothing is streamed. Your answers change nothing but the verdict text in front of you, and they last exactly as long as this tab does.