Best Codecs:
Speech (single speaker):     TOOLVOX-RT29HQ/3kbit/Mono --   375bytes/sec =  1.4 meg/hour
Speech (multiple speakers):  VQF 11kHz/8kbit/Mono      --  1000bytes/sec =  3.6 meg/hour
Music/Speech (nice quality): VQF 22kHz/20kbit/Mono     --  2500bytes/sec =  9   meg/hour
Music (near CD-quality):     VQF 44kHz/80kbit/stereo   -- 10000bytes/sec = 36   meg/hour

Audio Tools (in order of excellence)

Toolvox provides the highest compression for speech:

If you like to listen to other very interesting stuff in toolvox format click here and follow the audio links.

SoundVQ/ TwinVQ offers the best compression for high quality audio:

TwinVQ/SoundVQ is not in widespread use yet, but it simply is the best compression.
MP3 is also worthwhile, mainly because one can decode the file back to a WAV file. This may be useful in order to compress it differently, once an even better compression comes about.
Cool is a very good tool to edit your sounds and also to save in Real-Audio.
ToolVox is the preferred compression if you just need the words to be understood.
It takes almost any wav-file and encodes fairly fast, and make incredibly small files: example (22k sample rate, 20kbit/sec)

TwinVQ/SoundVQ is the best allround compression. You may need to save the WAV file as a 11kHz sample rate to get the best compression.
SoundVQ rates:
input 11kHz  output  8kbit or 10kbit  per channel
input 22kHz  output 20kbit or 24kbit  per channel
input 44kHz  output 40kbit or 48kbit  per channel

Links: WWW Audio Compression FAQ
Realtime speech and voice on the Internet

    



back to the index 1