Why does my iPhone recording get louder and quieter on its own?

This is a specific, recognisable pattern once you know what to listen for: the room tone or background hum noticeably rises in the gaps between sentences and drops back when speech resumes, almost like the recording is breathing. It is a different complaint from a recording that is noisy all the way through, and it has a different, fairly mechanical cause.

What is actually happening

A phone microphone does not simply capture sound at one fixed level and hand it over. Underneath, the system is running automatic gain control — a continuous, moving adjustment that turns the input up when it is picking up very little and eases it back when it is picking up plenty. The goal is sensible on its own terms: keep a normal conversational voice sitting in a usable range without a manual level dial for you to fuss with.

The side effect shows up the moment you stop talking. With no voice to key off, the only signal left is the room — a fan, traffic, the general hiss of the space — and gain control turns the sensitivity up to keep something audible in a level it expects speech to occupy. That is what makes a fairly quiet room tone swell for a second or two of silence. The instant you start talking again, the voice is loud enough on its own, so the gain eases back down and the swell disappears into it.

None of this is something an app decides. It happens at the input stage, in the same place clipping happens, well before any recording setting or transcription step gets involved.

Why it is not the same as ordinary background noise

A room with a constant fan running sounds noisy throughout a recording, at roughly the same level whether you are mid-sentence or silent — that is a placement and room problem, and the fix is distance and a quieter room. What is described here is different: the noise is not constant, it specifically pulses in the pauses, in a pattern that tracks your speech rather than the room. Confusing the two sends you looking for a quieter room when the room may already be reasonably quiet — the pulsing is coming from how the gain reacts to silence, not from how much noise is actually present.

It is also a different failure from a recording that sounds distorted. Distortion is a peak clipped flat on a specific loud moment; this is the noise floor drifting up and down between sentences. One is about too much signal arriving at once, the other is about how the system fills in the gaps when there is almost no signal at all.

What makes it more or less noticeable

Steady, continuous noise is what gets exaggerated most

Gain control reacts to an average level over a short window, so a steady hum — air conditioning, a computer fan, road noise through a window — is exactly the kind of sound it ends up amplifying during a pause. An irregular sound, like an occasional door closing, does not pulse the same way, because it is not there long enough for the gain to settle around it.

How close the voice is to the microphone

The wider the gap between how loud your voice is and how loud the room is, the less the gain has to move between the two. Speaking close to the phone keeps your voice comfortably above the room, so the adjustment needed during a pause is smaller and less audible. A voice picked up faintly from across a room is much closer to the room's own level already, which gives the gain further to travel and makes the swell more obvious.

How quiet the room actually is to begin with

A genuinely silent room has very little for the gain to amplify, so turning up during a pause turns up next to nothing. This is why the effect is most noticeable in rooms that sound fine to the ear at the time — a light, steady hum you had already tuned out — rather than in rooms that are obviously loud throughout.

What does not fix it

The recording quality setting has no bearing on this. Standard, high and ultra change the sample rate the signal is measured at — how finely a given level of loudness is stored — not how the input decides what that level should be from moment to moment. A pumping room tone at standard quality pumps exactly the same way at ultra; a finer sample rate just stores the same rising-and-falling level in smaller time slices. Mono versus stereo is unrelated for the same reason: neither one touches how gain is calculated.

What actually helps

• Deal with steady noise at the source before it needs amplifying: turn off the fan, close the window, step away from the air conditioner. Removing what the gain would swell is more reliable than anything after the fact.
• Get closer to the microphone. Widening the gap between your voice and the room reduces how far the gain has to travel between speech and silence, which is the direct cause of the pulsing.
• Do not expect a quality tier to help. If a setting claimed to fix this, it would be claiming control over something the sample rate has never governed.

None of this requires anything close to a silent room. It needs the voice to sit clearly above whatever the room is contributing, which is the same lever that helps almost every recording problem that starts with "why does this sound off."

Hearing it before it is a whole recording

The pattern is easiest to notice while it is happening rather than on playback afterwards, because in the room your own hearing does the same kind of adaptation and tunes the swell out. Listening to the actual input as a recording runs — rather than trusting your ear in the room, which is doing you a disservice here — is the practical way to catch a noisy spot before an hour of audio has the same pulse running through every pause.

How Voice Studio fits into this

Voice Studio does not put a manual gain dial in front of you — the input level is handled by the system the same way it is in most phone recorders, and the quality tier you choose, standard at 22.05 kHz, high at 44.1 kHz or ultra at 48 kHz, all mono AAC, governs sample rate rather than how gain is calculated, for the reasons above. Pro's live monitoring lets you listen to the input as it happens, which is the direct way to notice a pulsing room tone while you can still move the phone or deal with the noise at its source rather than discovering it on playback. Noise reduction, bundled with it, works on suppressing steady background noise generally rather than something built specifically around the gain swell — but a quieter noise floor to begin with gives the automatic leveling less to exaggerate between sentences in the first place.

Common questions

Is this the same thing as background noise in a recording?

Related, but not the same. Ordinary background noise sits at roughly the same level throughout. This is that noise specifically swelling during pauses and settling back down once speech resumes, which is a sign of automatic gain control reacting to silence, not of the room itself changing.

Does a higher quality setting fix the pulsing?

No. Standard, high and ultra change the sample rate the audio is captured at, not how the system decides gain from moment to moment. The pattern is identical at every quality tier.

Why do I only notice it in some rooms and not others?

It is most noticeable with a steady, continuous sound — a fan, air conditioning, road noise — because gain control reacts to that kind of average level. A quiet room with only occasional, irregular sounds rarely shows the pattern, because there is nothing steady for the gain to settle around during a pause.

Does live monitoring actually help with this?

Yes, more than most settings would. The swell is easiest to catch in the moment, before your own hearing tunes it out the way it does in the room — listening to the input as it happens is the practical way to notice it and deal with the noise at its source.