Subject: Re: [linux-audio-user] jaming over the internet?
From: Juhan Leemet (juhan_AT_logicognosis.com)
Date: Wed Oct 01 2003 - 21:51:31 EEST
GuyCLO~ wrote:
>On Mon, 29 Sep 2003 16:58:22 -0230, Juhan Leemet wrote / a écrit:
>> Now add in the long
>> echo delay, and I suspect that's too much to handle: 3 time bases: what you
>> want to play, what you are playing (feel?), and what you hear. Comments?
>
>I have read your interesting text, and I think the points you bring are
>correct.
>But I think that the idea should not be dismissed just for the reason that
>it's not suitable for some people/sorts of music. Some people may want to
>play new-age in the same basement/city over IP :)
>In some cases, this app will be very valuable. Is it really hard to use
>aconnect, as Gustavo suggested?
Sorry. It is not my intention to dismiss the idea, but try to understand and
discuss the limits/boundaries. After all, we are constrained by physical
laws. If the (effective) transmission time from L.A. to NewYork is greater
than some perceptual delay (I don't know enough, to say what that is), then
the technique is not going to work (between L.A. and NewYork, but maybe it
will work between L.A. and Santa Monica?), at least for streaming
high-bandwidth audio. Popular press often gets this stuff wrong, since they
prefer sensationalism. No point attacking a problem that can't be solved. In
the same city, I guess it could be feasible, maybe, but we'll have to try it.
I recall the hoopla about video conferencing, and my computer consulting
client (a telco) basically gave up (at least for their own development
projects) after a couple of years of trying. Calculations all seemed good.
Marketing guys were enthusiastically (over) selling, as usual. In practice,
just didn't work well enough. This client was a telco, so they had available
all the bandwidth they wanted! Delays, jitter, quality, "artifacts" (of
compression?), and just plain unreliability made it horrible. This was for 2
and 3 way tele-conferencing. I think one-way video streaming (such as some
instructional videos) does (sort of) work, and mostly because you can add as
much delay as you want to allow filtering, buffers to catch up, etc.. We are
all being "trained" to accept lower quality video, and jerkier motion, etc.
Hopefully we won't degrade our music into those lame MIDI demos of "blues" or
"jazz". Reminds me of some classical music professors I have heard trying to
form a jazz band: technically great, mathematically precise, theoreticaly
"faultless", but no soul, and therefore lame and uninteresting. Not trying to
slag classical music professors. I imagine there are some here. "In theory
there is no difference between theory and practice. In practice there is." As
we compress and process, we remove some soul and feel. A lot of contemporary
music seems "over-processed", and therefore sounds all much the same. Maybe
the fault is mine, and I'm being far too demanding (discriminating?).
Internet jamming is easier with MIDI traffic. I had not thought in those
terms, probably because I'm currently (re)learning some guitar, which does
not translate into MIDI very well (expensive rigs, tracking is a problem). My
conception was that jamming = (two-way) "real-time audio streaming" (hard!).
Steve mentioned that some universities have special high-bandwidth,
low-latency interconnections set up. They must be quite special, and I'm not
sure how easily those capabiliites will be available to the rest of us.
I am finding this information interesting. Food for thought (and experiment).
I should get back to wrestling with my M-Audio Audiophile 2496, and getting
it to work with various applications. Too many distractions.
-- Juhan Leemet Logicognosis, Inc.
This archive was generated by hypermail 2b28 : Wed Oct 01 2003 - 22:28:17 EEST