If you've ever heard a recording of your voice – in a video, for example, or on somebody's voicemail – you know that the sound you hear in your head when you speak is different from the one heard by everyone else. Why is that, and why do we tend to prefer the voice we hear in our heads?
Hank Green answers the first half of this question in the latest installment of SciShow, and hints at the answer to the second:
Our voices that we hear through our own heads actually sound deeper and more resonant... when you hear it recorded, you're hearing it as other people hear it. This doesn't mean that it's worse, it just feels really odd to hear your voice without all the components you're used to, and that can be unsettling since... you've been listening to yourself for, I presume, a pretty long time.
Emphasis mine, because it explains why we tend to prefer the sound of our voices not as they sound to everyone else, but as they sound to us, vibrating around inside our heads: Basically, the version of our voice we hear most often is the version that we grow to prefer. Psychologists call this the mere-exposure effect. It's a phenomenon that's been demonstrated with an array of stimuli (words, paintings, sounds) and across cultures. It may also explain why many people think they look weird in photographs.