Why does your voice sound wrong in a recording?

Two signals at once

When you speak, sound reaches your own inner ear by two very different routes simultaneously. This surprises most people, because hearing tends to feel like a single, unified process. But the two routes carry different information, and your brain blends them into one seamless perception. Nobody tells you this is happening — which means nobody tells you that the voice you hear in your head is a composite, and that only half of it is available to anyone else.

The first route is air conduction: the familiar one. Your vocal cords vibrate, producing sound waves that radiate outward into the air, travel into your ear canal, and set your eardrum in motion. This is the same path that carries all external sounds to you, and it's the only path available to anyone standing nearby — or to a microphone.

The second route is bone conduction: the hidden one. Vibrations from your vocal cords also travel directly through the bones and soft tissue of your skull to the cochlea — the fluid-filled structure in your inner ear that converts mechanical vibration into the signals your brain reads as sound. This route bypasses the air entirely. It's a shortcut that exists only for the person doing the speaking.

	Air conduction	Bone conduction
How it travels	Sound waves through the air and into the ear canal	Vibration through the skull directly to the cochlea
Who hears it	Everyone nearby, including microphones	Only you, while you're speaking
Frequency profile	Full range, filtered by the air medium	Emphasises lower frequencies more efficiently
In recordings	Yes — this is all a microphone captures	No — lost entirely once the mic is involved

The private bass track

The two routes don't carry identical information. Bone conduction is particularly efficient at transmitting low frequencies — the deeper, richer components of sound. So when you hear your own voice during speech, you're receiving a version that has been boosted at the low end, courtesy of the bone route. The result is a voice that sounds fuller and deeper to you than it does to anyone else in the room.

A microphone captures only what's in the air. So a recording cuts the bone conduction signal out entirely, leaving you with the air-only version — thinner, higher, and stripped of the private bass track that existed only inside your skull.

Imagine listening to a live concert from the stage rather than from the audience. The musicians hear the same music the audience hears, plus the resonance of the floor beneath their feet and the vibration of their instruments through the wood of the stage. Remove that floor, and the sound changes — but what the audience hears is exactly what was always reaching them. The recording is the view from the audience. Your internal sense was always the view from the stage.

Proof you can hold in your hand

Bone conduction headphones make this principle concrete. Rather than sending sound through the air into the ear canal, they press small transducers against the skull and deliver vibration directly to the bone. They work precisely because the pathway is real, reliable, and entirely distinct from air conduction.

Audiologists have used the difference between the two pathways as a diagnostic tool for decades. By testing how well a person hears sounds delivered via air versus via bone, clinicians can pinpoint where in the hearing system a problem originates. Someone who hears bone-conducted sound normally but struggles with air-conducted sound likely has an issue in the outer or middle ear. Someone who struggles with both has something affecting the inner ear or the auditory nerve. The two routes are genuinely separate enough to be tested independently — and that independence is exactly what creates the gap in how you perceive your own voice.

What professionals learn from this

Singers, broadcasters, and voice actors all go through a version of the same adjustment. Early in training, hearing yourself through a microphone and playback feels disconcerting in exactly the way it does for everyone else. The internal sense insists: that's not what I sound like. But the internal sense is structurally unreliable. Training involves learning to trust the recording over the internal impression — because the recording is what the audience hears, and the internal impression never was.

The practical upshot for everyone else is simpler: the discomfort fades with exposure. People who record themselves regularly — for podcasts, presentations, voice notes, video calls — report that the strangeness recedes within weeks. The brain has some capacity to recalibrate. It doesn't eliminate the gap between the two signals, but it learns to accept the recorded voice as a valid representation of the self rather than a distortion of it.