Will your speaker hold up at Saturday’s party? Play our test track, set your phone about 6 feet away and turn the volume up step by step. The test watches for distortion and estimates whether you have a speaker loud enough for a party, from a desk to a backyard.
This party speaker test tells you how far your speaker gets before it distorts, from a desk to a backyard, by measuring it with your phone about 2 meters (6 feet) away while you turn the volume up. It's a quick way to find out whether you have a speaker loud enough for a party, from a small Bluetooth party speaker to a big battery-powered one, because it measures the one you own in the place you'll use it instead of trusting the watts on the spec sheet. The test track leaves the 300 Hz to 3 kHz band empty, and energy that shows up there as you turn it up is almost always distortion from your gear.
It's an estimate built on relative levels. A phone microphone isn't calibrated, so the test measures how far your speaker rises above the room's noise floor, and the scenarios are thresholds we set. Use it to compare speakers and to find how loud yours can go before it distorts.
How to tell if you have a speaker loud enough for a party
You need the speaker, a phone and a couple of quiet minutes.
- Pick your setup. With one device, the speaker is connected to your phone or laptop by Bluetooth or cable, and that same device measures with its microphone. With two, one plays the track and the other measures, and you always start on the one that measures.
- Put the meter about 2 meters (6 feet) from the speaker, out of its case and with the microphone uncovered. Keep everyone quiet for the second and a half it spends measuring the noise floor.
- The track starts low and fades up on its own over 10 seconds. After that, raise the system volume one step every two or three seconds. The margin and the scenarios light up as the test confirms the sound is still clean.
- When "On the edge" or "Distorting!" appears, back off one step and press Finish.
- Read the scenario. To compare against another speaker or another spot, repeat in the same place with the same background noise.
How the test spots distortion
It watches the gap in the track. The kick and bass sit below 150 Hz and the hi-hats above 5 kHz, with nothing between 300 Hz and 3 kHz. When an amplifier or a speaker runs out of headroom, it clips the waveform, and clipping creates harmonics that land right in that empty band. The test compares the energy in the gap with the energy in the music and warns you when the gap starts filling up.
It never decides on a single reading. It looks at the last loop of the pattern, about 2 seconds, subtracts the noise that was already in the gap before the track started, and compares the result against fixed thresholds or against the cleanest reading it has seen from your speaker, whichever is higher, because a small speaker with little bass starts out with a fuller gap without distorting. A reading only counts toward your score once the 2 seconds after it come out clean as well, and readings from a dirty stretch are thrown out even if you turn it down afterward, so the score stops climbing shortly after clipping begins.
We ran the detector through the same simulation with a fixed volume and pink room noise at three levels, 20, 30 and 40 dB below the track. The ceiling is the highest level the amplifier delivers without clipping, and the table shows what the test reports depending on how far past it the signal goes.
| Above the ceiling | Samples clipped | What the test shows |
|---|---|---|
| +1 dB | 0.3% | Clean |
| +2 dB | 0.6% | "On the edge," except in the noisiest room, where 9 in 10 readings stay clean |
| +3 dB | 0.9% | "On the edge" or "Distorting!" depending on the room |
| +4 dB | 1.4% | "Distorting!" |
| +6 dB or more | 13% or more | "Distorting!" |
With the volume rising 1 dB per second, we tried two kinds of noise, a regular and a small speaker, 30 to 120 readings per second (it depends on the screen) and rooms of different noisiness, with clipping starting at about 16 to 48 dB of margin. With hard clipping, across all 32 cases the final score landed between 0.5 dB below and 3.5 dB above the margin at which the signal starts to clip, with no warning before it, and the biggest overshoot came with the lowest margin. With soft clipping the score landed between 5.6 dB below and 3.7 dB above, because with plenty of margin it warns early and with little margin, late. If you turn the volume down after the warning, the score stays within 1.5 dB of what it would have been had you never gone past the ceiling. Below about 16 dB of margin (a very small speaker or a noisy room), the warning comes late or not at all and the score can overshoot by up to about 6 dB, so "Desk" and the low end of "Bedroom" are only an estimate of the margin. A real speaker also adds its own room and quirks, so read the result as an indication.
What each scenario means
Each scenario is a band of clean margin, the difference between the track's level and the noise floor the test measured when it started. The bands are our own call.
| Scenario | Clean margin over the noise floor | What it's good for |
|---|---|---|
| Desk | 12 to 15 dB (below that, "No result") | Background music at your desk or in a small room |
| Bedroom | 15 to 25 dB | Filling a closed room at a comfortable level |
| Large living room | 25 to 33 dB | A large living room with people chatting over it |
| Indoor party | 33 to 40 dB | Music front and center with people dancing |
| Backyard / outdoors | 40 dB or more | A backyard party or an open terrace |
The margin depends on how noisy the room is. A quiet room gives a higher number than one with a fridge, traffic or air conditioning in the background, so only compare results taken in the same spot. It isn't an SPL reading like a sound level meter's, either. And since it's measured at 2 meters (6 feet), outdoors, with no walls to bounce the sound back, every doubling of distance costs you 6 dB. That's why the backyard needs more margin than the living room.
What the test tells you for each kind of party
A dinner or a birthday at home
Background music doesn't take much. If your speaker reaches "Bedroom" or "Large living room" cleanly, that's plenty, since the point is for people to talk without raising their voices. A normal voice runs about 62 dB at one meter, according to the ANSI S3.5 speech intelligibility standard, and a raised one about 68. If the music forces people from one to the other, it's too loud for dinner.
A small Bluetooth speaker is usually enough here, even one the test leaves at "Bedroom". Placement changes the result too. Pushed against a wall or into a corner, the bass gets louder, while in the middle of the table, surrounded by plates and people, the sound spreads more evenly. Put the speaker where it will sit during dinner before you measure.
If the test gives you "Desk", check first that both the phone volume and the speaker volume are maxed, since they're two separate controls, then run it again. If it stays there, it may still do for a dinner of six or eight, just not loud enough to hear from the kitchen.
A house party with dancing
For dancing, aim for "Indoor party" or higher. With people talking loudly, which ANSI S3.5 puts at about 75 dB at one meter, the music has to sit above them, and the test measures an empty room, without the crowd that later soaks up sound and adds noise of its own. A speaker that barely makes "Indoor party" in an empty room will most likely fall short at midnight, so leave one scenario of headroom.
If yours tops out at "Large living room", you have three fixes. A second, identical speaker linked to the first adds about 3 dB (more on that below). Moving it closer to the dance floor helps too, since halving the distance gains you up to 6 dB. And for a big crowd, what you need is a PA speaker, which is far more sensitive than a home speaker; our guide to how many watts a party really takes has the numbers by crowd size.
Don't overdo it. Safe listening time shrinks fast above 85 dB, and our guide on how many decibels damage your hearing lists the times.
At an outdoor party in the backyard or by the pool
A speaker for an outdoor party needs more margin than one indoors, because outside there are no walls to send the sound back. Every time you double the distance, the level drops 6 dB, so at 8 meters (26 feet) it's 12 dB quieter than at the 2 meters (6 feet) where the test measures. That's why the backyard asks for more margin, 40 dB versus 33, a threshold we set.
Run the test outside, where the party will be, and without wind. Wind over the phone's microphone sounds like a low rumble, and depending on when it blows it can lower or inflate your result, or light up the warning. If there's traffic or a neighbor playing music, the margin will come out lower too, and that's exactly what you'll have to rise above that night.
Outdoors, the neighbors hear more than they do through your walls. Local noise ordinances set limits, especially at night, and to know how many decibels reach their window you need an absolute measurement, which this test doesn't give you.
With two linked speakers
Two identical speakers playing the same thing add about 3 dB, a little more in the bass if they sit close together. That's less than it sounds, because it takes about 10 dB for something to sound twice as loud, but it can be exactly the jump from one scenario to the next.
To check it, measure one speaker on its own first and note the margin. Then link them with your brand's party or pairing feature so they play together, and run the test again from the same spot, with the meter about 2 meters (6 feet) from both. If the margin doesn't go up by roughly 3 dB, look at placement or the link mode. In stereo mode each speaker plays one channel, but the test track is identical on both, so that works too.
Keep them close and aimed at the same area. Placed in two corners of a backyard, each one covers its own side and the margin at any single spot barely changes, even if the party as a whole sounds better spread out.
When the warning goes off right away or never at all
If it goes off at a low volume, look for something filling the gap that isn't your speaker. Voices fall right between 300 Hz and 3 kHz, so someone talking, the TV or the radio will trigger it; run it again in silence. An object rattling nearby, like a lamp or a pane of glass, is real distortion, just not from the speaker, and it's worth hearing before the party. Bass boost modes on many Bluetooth speakers also bring clipping on sooner, so try with and without and keep whichever goes further cleanly.
If it never goes off, your speaker may simply handle more than you're asking, especially if you haven't maxed out both volumes. If the margin stops climbing before you reach the top, even as you keep turning it up, the speaker most likely has a limiter, which holds the peaks down before the amplifier clips. In that case the test finds the limiter's ceiling, which is also its clean maximum.
In single-device mode, if the Bluetooth speaker has a microphone for calls, opening the mic can switch the link to hands-free mode, which sounds much worse. If the sound gets worse when the test starts, or the margin doesn't rise even though the speaker keeps getting louder, try two devices, one playing and one measuring.
Frequently asked questions
How loud should a party speaker be?
For a dinner, below conversation; if people have to raise their voices to hear each other, it's too loud. For dancing, much louder, and in our wattage guide we aim for an average of 95 to 100 dB on the dance floor. Don't go past that. The WHO asks venues and events to keep the average under 100 dB over any 15 minutes, and at that level NIOSH gives your ears about 15 minutes a day.
What is the best speaker for a party?
The one that reaches your party's scenario cleanly in the place you'll hold it. The watts on the spec sheet won't tell you, and two speakers with the same figure can end up several scenarios apart. If you're choosing between two, run this test on both in the same room, at the same distance and with the same background noise, and keep the one that goes furthest without distorting.
Is a 200W speaker loud?
The watts on a Bluetooth or party speaker rarely tell you how loud it gets. They're usually the total of its built-in amplifiers, sometimes a peak figure, and often with no word on the distortion they were measured at. Two 200 W speakers can sound very different depending on their sensitivity and their limiter. What actually tells you is the maximum level in dB, which few brands publish, or a test like this one in your own room.
How do you know how loud a speaker is?
This test tells you how far it goes cleanly in your room, as a margin over the noise floor and a scenario. For a decibel figure you need a sound level meter, and a phone one is only indicative, off by as much as 10 dB unless you calibrate it. A maximum SPL on the spec sheet tells you how loud it gets at one meter under the maker's conditions, not in your living room.
How does it detect distortion without lab gear?
With a track that leaves the 300 Hz to 3 kHz band empty. Clipping fills that gap with harmonics, and the test watches it over the last loop of the pattern instead of trusting a single reading. In our simulation the warning goes off with about 2 to 4 dB of clipping, a little later in a noisy room. The track is generated in your browser, so it's royalty-free.
Why scenarios instead of watts or decibels?
Because a browser can't measure absolute decibels with an uncalibrated phone mic, and "20 watts" doesn't tell you whether the backyard party will work. The test measures your speaker's clean margin over the room's noise at about 2 meters (6 feet) and translates it into five scenarios, with thresholds we set. It's an estimate, and that's how to read it.
What if it never shows distortion?
Your speaker may have headroom to spare at that volume, a limiter that holds the peaks down before clipping, or you may not have turned it up far enough. With a limiter, the test finds its ceiling, which is also its clean maximum. Otherwise, keep turning it up until the warning appears or it starts to sound bad.
Can I damage my speaker with this test?
A few seconds of distortion while you look for the limit won't hurt a healthy speaker, but don't leave it playing distorted. Sustained clipping is what heats up and burns out tweeters. When the warning lights up, back the volume off a step.