Human Voice – A Unique Identifier

It is uncanny how we can recognize an individual’s voice from a million other voices. A person’s voice is as unique as a fingerprint. In fact, no two voices on the planet are identical.

Voice characteristics are dependent on a number of anatomical factors, including: size and strength of the lungs, dimension of the airways (trachea), length and tension of vocal folds (cords), size and shape of the oral and nasal cavities and the relative positioning of lips tongue and soft and hard palates, in addition to the size of the torso and head and density of bones and tissues.

Even if two people shared identical physical characteristics, there would still be recognizable differences in voice due to differences in habitual movements of the articulators. No two individuals modulate their lips, tongue, palate, vocal cords or lungs in the same way when speaking.

Human voice on average encompasses a frequency (pitch) range from 125 to 8000Hz, and includes sounds ranging from 10 to 60dB in loudness when speaking at a conversational volume. Individual elements of speech (phonemes) occupy different regions of pitch and loudness on the speech spectrum. Consonants are typically softer and higher in pitch than vowels. The softest and highest consonants of all are the unvoiced sibilants /f/, /sh/, /th/, /s/, /ch/, which do not involve vocal fold vibration, but are formed purely by the friction from air escaping around the mouth, teeth and tongue.

In terms of fundamental pitch, there is considerable overlap between male and female voices. On average women have a first harmonic at 210Hz, with men’s occurring at 130Hz. Voice is of course comprised of numerous overtones and temporal characteristics, giving each person a unique timbre. Criminalists may use spectrographic analysis to identify a suspect on the basis of their timbre.

Interestingly, it is harder to pick an individual’s voice over the phone. Telephony transmits only a restricted range of frequency, thereby removing important auditory cues occurring below 300Hz and above 3400Hz.

Hearing loss tends to progress in the high frequencies first, resulting in a gradual loss of consonant detection. Such high tone hearing loss causes speech to become muffled and unclear, despite little effect on overall volume. As hearing loss progresses, it typically affects lower regions of pitch, making more and more elements of speech inaudible to the listener. Eventually speech may become incomprehensible.