Editorial - Defining HD voice

HD voice as a term is coming close to becoming like unified communications — defined and stretched and abused by marketing and PR people to mean exactly what is necessary for the latest product and press releases.

Over the past eight months, I have heard a number of definitions for HD voice.  Let me codify and clarify what I — and nearly everyone else who has used the term — means when they use the phrase “HD voice” —

1) A typical PSTN call delivers audio in the range of 300 to 3400 hertz — a range of 3.4 kHz typically.
2) The baseline for wideband voice, G.722,  delivers audio in the range of 30 to 7000 hertz – so there’s a range of 7 kHz
3)  This would make G.722 “twice as good”/2x in terms of hertz as a typical PSTN call.
4) THEREFORE, HD voice is defined as the ability to deliver audio at least as twice as good as a PSTN phone call.
5) FURTHER, any discussion of HD voice should include the codec or codecs supported, since there are numerous codecs that are described as “wideband” or
“superwideband” or other terms.  Other codecs may deliver up to 5.88 times (or 6, if you round upward) as much audio as a typical PSTN call.
6) While some may use octaves to describe the range of sound being delivered on a phone call and this discussion can be somewhat refreshing,  many/most telephony-heads and ordinary mortals not steeped in multiple years of choir practice do not think in terms of singing, four octave or five octave voices, or the fact that a piano covers 7.5 octaves.
For reference/discussion purposes —
  • The human voice ranges from 80 hertz to 14,000 hertz.
  • Humans hear in a range from 20Hz to 20,000Hz.
  • A PSTN call can deliver much less than 3.4 kHz
  • A dog can hear from 40 Hz to 60,000 Hz.
Be Sociable, Share!

Comments are closed.