Re: Pathetic audio at 28.8 on Mac

Tim_Dorcey@cornell.edu
Fri, 7 Apr 1995 12:02:51 -0400


At 12:26 PM 4/7/95 +0200, Jan Engvald LDC wrote:
>
>The delta-mod "16 kb/s" audio actually produces 21 kb/s data.

This is true with sampling set to 50 ms. Selecting 100 ms samples will
produce half as many audio packets (each with twice the payload) and cut
down on header overhead. With that setting, I observe 18-19 Kbps for the
16 Kbps encoding. Then, 28 bytes of IP + UDP header per packet x 10
packets/sec, adds another 2.1 Kbps to the total data rate, which should fit
comfortably in 28.8, even stepped down some.
Note that the next version of CU-SeeMe will include IP+UDP headers in
the displayed data transmission rate, and also will allow 1 Kbps increments
in the data transmission (and reception!) cap controls. We never imagined
folks would be running CU-SeeMe under conditions where those few bytes
could make a difference. Now that we know, there's some other stuff we
plan to do as well.

Tim
__________________________________________________________________
Tim Dorcey Tim_Dorcey@cornell.edu
Sr. Programmer/Analyst (607) 255-5715
Advanced Technologies & Planning
CIT Network Resources
Cornell University
Ithaca, NY 14850
__________________________________________________________________