Open Source Codec Encodes Voice Into Only 700 Bits Per Second (rowetel.com)
Longtime Slashdot reader Bruce Perens writes: David Rowe VK5DGR has been working on ultra-low-bandwidth digital voice codecs for years, and his latest quest has been to come up with a digital codec that would compete well with single-sideband modulation used by ham contesters to score the longest-distance communications using HF radio. A new codec records clear, but not hi-fi, voice in 700 bits per second -- that's 88 bytes per second. Connected to an already-existing Open Source digital modem, it might beat SSB. Obviously there are other uses for recording voice at ultra-low-bandwidth. Many smartphones could record your voice for your entire life using their existing storage. A single IP packet could carry 15 seconds of speech. Ultra-low-bandwidth codecs don't help conventional VoIP, though. The payload size for low-latency voice is only a few bytes, and the packet overhead will be at least 10 times that size.
I wonder how it performs on tonal languages like Cantonese.
http://michaelsmith.id.au
Encoding voice more efficiently has implications far exceeding the amount of storage space required to save it. There's a reason why the article is comparing the new codec to single sideband. When transmitting digital data over radio, it pretty much invariably (nowadays) means some sort of spread spectrum transmission. The fewer bits required per second means the less spectrum you are having to spread your signal over, this the more concentrated your signal is. A radio transmitter has a fixed power output, so if you are smearing that power over less band, then you have a stronger signal.
It is a testament to the amateur radio pioneers of the past that an analog radio transmission mode invented over a hundred years ago is, just now, being possibly rivaled in its efficiency.
Actually, our modems degrade gracefully. The least-protected bits go wrong with low bit-error rates, and the more protected bits survive. It takes a high bit error rate to kill it. So bit errors result in the speech being "off" but not dropping out.
Bruce Perens.