You're previewing the Bennett demo family. Numbers, schedule, and kids are sample data.Start your own
Loading page…
You hear yourself in a video and think: that is not what I sound like. Everyone else in the room thinks it sounds exactly like you. One of you is hearing something the other one isn't.
Nearly everybody hates hearing themselves recorded. It is such a reliable reaction that psychologists gave it a name back in the 1960s: voice confrontation.
And the complaint is always the same shape. Not "that's a bad recording" — something much more specific. That doesn't sound like me. Higher than me. Thinner than me. A bit like somebody doing an impression of me, badly.
Meanwhile your friend, standing right there, listens to the same clip and doesn't blink. To her it sounds precisely like you, because it is precisely like you.
The obvious explanation is that the recording is at fault. Cheap microphone, phone speaker, compressed video — something in the chain mangled you on the way through.
There is a simple reason that can't be the answer.
If the machine had changed your voice, everyone would notice. Your friend would say "you sound odd on this" — the way she would if you'd recorded it underwater. She doesn't. Play a clip of you to twenty people who know you and every one of them names you in about half a second.
So whatever the difference is, it is a difference only you can hear. That rules out the microphone immediately, and it points somewhere much more interesting: at your own head.
Deep inside your head, past the eardrum, sits the part that actually turns vibration into something your brain can use. Everything you ever hear has to get there. There are two ways in.
Route one is through the air. Your mouth makes a sound, it travels out into the room, some of it curls back round to your ear canal, wobbles your eardrum, and in it goes. This is the route every other sound in the world takes, and it is the only route anyone else has to your voice.
Route two goes through you. When you speak, your vocal folds are vibrating a few centimetres from your ear, inside your own head. That vibration travels straight through the bone and tissue of your skull to the same destination. No air involved. It is called bone conduction, and it is happening every time you talk.
So you hear your own voice on both routes at once, mixed together. Everyone else gets route one only.
Now the crucial part. The two routes are not the same in what they carry.
Bone is heavy, dense stuff, and it passes the low notes along far more generously than the high ones. So route two arrives loaded with the deepest parts of your voice — the warm, resonant bottom end.
Add that to route one, and the voice in your head is lower, fuller and richer than the one leaving your mouth. It has extra bass welded on that nobody else has ever received.
A recording captures route one. Only route one. That is not the machine subtracting something from your voice — it is the machine failing to add something that was never really in your voice to begin with.
The recording is not a distorted version of the voice in your head. The voice in your head is your real voice plus your skull. The recording is the one everybody else has been listening to all along.
You can catch bone conduction in the act in about three seconds.
Press your fingers firmly into your ears, blocking them as completely as you can. Now hum.
Your hum gets louder. Noticeably, obviously louder — and deeper too.
Stop and think about how odd that is. You have just blocked your ears. Everything else in the room went quieter: the fan, the traffic, somebody talking to you. Only your own hum went up.
Hold the hum steady and take your fingers in and out a few times, and you can hear it happen exactly as your fingers land. Then, without changing anything else, get somebody near you to hum instead. Theirs behaves like everything else — fingers in, quieter.
Blocking your ears makes every sound in the room quieter — except your own humming, which gets louder. What is the best explanation?
Here is a fact worth sitting with. For almost the whole of human existence, not one person ever heard their own voice from outside their own head. Put these in order.
Tap whichever you think happened next
Think about what that means for everyone before 1877. Not a single one of them ever found out what they sounded like.
Every singer, every teacher, every person who ever told a story to a room — all of them went their whole lives hearing only the skull version, and never once heard the version their friends heard.
You get to. It costs you about four seconds and one small cringe, and that cringe is the sound of you meeting a version of yourself that everybody else already knew.
You insist the recording sounds nothing like you. Your friend insists it sounds exactly like you. Who is hearing the version of your voice that other people actually deal with?
You’re previewing as a parent — nothing here is recorded.
Self-heard voice combines air conduction with bone conduction: vibration from the vocal folds travels through the skull directly to the cochlea as well as out through the air. Bone conduction transmits the lower frequencies far more efficiently than the higher ones, so a talker's own voice sounds lower and fuller to them than to a listener. (Standard audiology; von Békésy's classic work on bone conduction, and Reinfeldt/Stenfelt et al., 'Hearing one's own voice during phoneme vocalisation', J. Acoust. Soc. Am., on the relative contribution of the two paths.) The occlusion effect: blocking the ear canal raises the perceived loudness of one's own voice, particularly at low frequencies, because bone-conducted sound that would normally escape from the open canal is trapped. This is the everyday 'hum with your fingers in your ears' demonstration and is routinely used in hearing-aid fitting. (Standard audiology.) 'Voice confrontation' — the discomfort people report on hearing a recording of themselves — was named and studied by Holzman & Rousey, 'The voice as a percept', Journal of Personality and Social Psychology 4(1):79-86 (1966). Édouard-Léon Scott de Martinville patented the phonautograph in 1857; it inscribed sound waves onto lamp-blackened paper and had no playback mechanism. His 9 April 1860 phonautogram of 'Au Clair de la Lune' is the earliest recognisable recording of a human voice. It was first converted to audible sound by the First Sounds collaborative, using optical scanning, in March 2008 — 148 years later. (First Sounds; Library of Congress; Smithsonian.) Thomas Edison's phonograph, 1877, was the first device able both to record sound and to reproduce it, making it the first point at which anyone could hear their own voice by air conduction alone. No medical content: this episode describes ordinary hearing and gives no advice or assessment of anyone's hearing.