At 09:20 AM 11/19/2004 -0500, andy kolovos wrote:

>Folks,
>
>I realize that the size of an .mp3 file will have an influence on the time
>it takes to download and play the file.  What effect does file size have on
>streaming--in particular streaming using a .m3u file?

I do not use streaming myself, so can only repeat what I've been told.
Since it seems to be logical as well, here 'tis.

In streaming, the bitrate issue is somewhat different from that for
download. In essence, your server has to be able to supply the concurrent
streams and the recipient's connection must be fast enough for one such. If
you assume that those on dialup have near-56 Kbps effective speed, then I
would recommend 32 Kbps (22.05 ksps if monaural) to drive them. That would
allow about 32 concurrent connections at full speed. Cutting back to 24
Kbps would reduce the load on the server but would not help the recipient.
24 is unlikely to be supported in practice on a 28 Kbps connection and
congestion should permit 32 on a 56 Kbps connection if it permits 24.

Note that at low rates, especially for speech, many find RealAudio to be
more acceptable than MP3.


Mike
--
[log in to unmask]
http://www.mrichter.com/