Test Your Microphone Before the Party: 5 Steps

Give it ten minutes, not thirty seconds. To test a microphone before a party, check three things in order: that the mic hears you cleanly, that the speaker reproduces what it is given, and how far behind your mouth the sound lands. Almost every mic failure in front of a room is one of those three, and all three are checkable in a browser before the first guest arrives.

Do it with the gear you will actually use, in the room you will stand in. A test at your desk with AirPods proves nothing about a phone held at arm's length in a crowded kitchen.

The ten-minute run-through

Five passes, roughly two minutes each. Nothing needs an install or an account. Only the last one needs a second person, and it is the most useful of the five.

Run them in that order. There is no point measuring delay on a link whose mic is already broken, or tuning a voice preset on a speaker that rattles. The same pass works to check a microphone before an event of any shape: a wedding toast, a quiz night, a school hall.

Step one: test the microphone before the party starts

Open the mic test on the phone you plan to use, grant microphone permission, and talk at it the way you will talk on the night. Three readings matter.

• Level. A normal speaking voice should move the meter confidently. If it barely twitches, you are too far from the phone or something is holding the input gain down.
• Peak. The peak reading should stop short of the top. Pinned at maximum means clipping, and clipping crackles through a PA no matter how good the speaker is.
• Sample rate and channel count. The tell nobody checks. A low reported sample rate with a single channel is a strong hint that a Bluetooth input has landed on the headset (HFP) profile rather than A2DP, and that you are about to broadcast the room through a telephone. Read it as a hint and not as proof: a plain wired mono microphone can report the same pair of numbers for reasons of its own, with no Bluetooth profile involved. Confirm it by switching the input back to the phone's built-in mic and watching whether the sample rate climbs.

Then use the five-second record-and-play-back. This is what people mean when they search for a karaoke mic test online: not a meter, but hearing yourself. Sibilance, plosives on hard consonants, a fridge hum you had stopped noticing. The recording stays in the browser on your own device.

Step two: check the speaker, not only the mic

Half of what gets called a bad microphone is a speaker being asked for something it cannot do. The speaker test runs left and right channel checks, a variable tone, a 20 Hz to 20 kHz sweep, and a 20 Hz to 200 Hz bass sweep.

Play the full sweep once and listen for the point where the low end stops existing. Every small enclosure has one. Then play the bass sweep at the volume you expect on the night. Buzz, rattle or port chuffing means the speaker is at its limit with nothing but a test tone, and a loud voice will push it past.

Run the left and right checks even with one speaker. Many portable units sum both channels into a single driver, which is fine, but with a stereo pair you want to know which cabinet is which before you place them. Read what causes two Bluetooth speakers to drift apart first: a pair out of step with itself is worse than one speaker alone.

Step three: measure the delay you will hear

This is the number that decides what your party can do. The latency test plays a 30 ms chirp between 1 and 6 kHz, records it back through the mic, cross-correlates the two, runs five times, takes the median, and subtracts the travel time for the distance you enter. That is your real round trip, not a spec sheet.

Published figures for the underlying links, so you can tell whether your reading is normal:

For what those numbers feel like: under about 20 ms is unnoticeable, 40 to 50 ms is where lips and sound visibly stop matching, and past about 100 ms it feels like a bad phone call. Monitoring your own live voice is far harsher than listening passively, which is why a delay nobody notices during a toast is unbearable when you sing. More in how much audio delay people actually notice.

Step four: walk the room for feedback

Feedback is a loop: the mic hears the speaker, that gets amplified, the speaker gets louder, repeat until the howl. Open the live monitor, set the gain where you expect to run it, and walk the route you will actually take. Toward the speaker, past it, into the corner, back to the doorway. Find where it rings before a room full of people finds it for you.

Three fixes, in order of how much they help: stand behind the speaker rather than in front of it, aim it at the audience rather than at yourself, and lower the gain. A step backwards beats any filter.

The Bluetooth Mic app reads the live output route and caps gain at 60 percent whenever that route is the phone's own built-in speaker, the configuration that howls instantly, then lifts the cap once a Bluetooth speaker, headphones or AirPlay takes the route. The longer version, including why the distance between mic and speaker matters more than the volume knob, is in the guide to microphone feedback with a Bluetooth speaker.

Step five: the soundcheck with a real voice

This is the soundcheck, and doing it with the phone itself is the point. Hand the phone to somebody else and go stand where the guests will stand. You cannot judge a PA from behind the microphone. Have them talk normally, then loudly, then turn away mid-sentence, because that is what everyone does.

Two things change between an empty room and a full one. Bodies absorb high frequencies, so speech loses clarity, and forty people talking raise the noise floor you have to clear. You will want more level than the empty room suggested, and level is what invites feedback, so leave headroom: speaker around three-quarters, app gain in the middle, nothing pinned at maximum.

If clarity is the problem rather than loudness, a filter beats a volume knob. The Megaphone preset high-passes everything below 600 Hz and pushes hard at 2.5 kHz. Consonant intelligibility does not sit at one frequency; it depends on a broad upper band, roughly 2 to 8 kHz for fricatives and sibilants, which is exactly what a narrowband phone-call connection throws away. Lifting the lower part of that band is enough to make a voice cut through crowd noise without getting louder. Bright lifts above 3.5 kHz more gently. Deep and Warm go the other way for a mellower announcement, and Radio and Robot exist for games. All eight, each with its real frequency response curve, are in the live voice effects tool.

What to check before you put the phone down

The audio can be right and the evening still fall over on housekeeping.

• Pair in iOS Settings, well before the party. No iPhone app pairs Bluetooth devices; there is no public API for it. An app plays into whatever route iOS has already chosen. Full sequence in connecting an iPhone microphone to a Bluetooth speaker.
• Do not use the speaker's own built-in microphone. A Bluetooth link carries one profile at a time, so asking for that mic drops stereo A2DP for the narrowband mono call profile, and iOS gives you no say in it. Phone mic in, speaker out sounds better.
• Turn on a Focus mode. An incoming call takes the audio session away from you in front of everybody.
• Charge everything and check the speaker's idle timeout. Plenty of portable speakers power down after a stretch of silence, which is exactly what precedes a speech.
• Keep a cable in your pocket. If the delay measures badly or the link keeps dropping, a wired run into the speaker's aux input sidesteps the codec entirely. The size of that gap is in wired versus Bluetooth microphone latency.