I'm going to go out on a limb here and guess Daniel is interested in speex
for real time audio conversations between players in a massively
multiplayer online game.
My bet is that there are very large differences between what a gamer would
consider as "good" and an audiophile would consider as
"good", based on a
long and glorious history of games short changing audio in most any
fashion in order to save CPU and/or development time -- with a few notable
exceptions, of course.
Gamers would probably consider the speech compression good if their
friends were still mostly intelligible -- assuming they were intelligible
to begin with. They would consider the bit rate too high if there was a
noticeable impact on the catch-all "lag", data transfer latencies on
real
time game data.
Consider that the gaming communication state of the art right now is for
the most part ASCII text, text->speech synth or prerecorded taunts. Simply
being able to hear your friends' real voices -- at all -- would be a huge
leap forward. The whole point is really to blow your friends into tiny
bits. The post-mortem epithets are mere icing on the cake. Even more
important is the gameplay win. You dont have to take your hands off the
control device in order to communicate while playing.
Given a 28 kbps uplink, one would perhaps wish for the most trivial impact
to bandwidth possible for a real time game, perhaps something on the order
of 2-4 kbps, and <10% cpu. The CPU load may be negotiable, based on
increasing reliance on GPU's and prevalent SIMD units for audio
processing.
Ian Ollmann
--- >8 ----
List archives: http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to
'speex-dev-request@xiph.org'
containing only the word 'unsubscribe' in the body. No subject is
needed.
Unsubscribe messages sent to the list will be ignored/filtered.