First, let me release the fruit of my greater than
mild enthusiasm about sfront (or mpeg-4sa or saol or sasl
or all of those, or what ever I have been using).
I was very well impressed by the performance of an
orchestra that used real-time audio in and out on
a moderately endowed Linux box. I put in a couple
of allpass's and a reverb, and I was able to use a latency
of .0001 with not one glitch. I am truly at home now.
Couple of questions:
1) By default, real-time performances seem to die after about
a minute or so. I was able at one time to keep a particular
performance alive by putting dummy events into the score.
Now I have a perf with -linux audio in and out, that I am having
problems keeping afloat. I just wondered what people normally
do to keep the program alive.
2) I was not able to get the -ain std option to work. I can use
my recording program to dump into a foo.wav file, and then use
-ain foo.wav for audio source, and all is well. Then I change
to -ain std and call the binary with sa < foo.wav or
cat foo.wav | sa, and this way, I can get no apparent input with
the same orc that worked with -ain foo.wav. Any ideas?
-ain linux works very well, and it is a boat-load of fun.
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 11:46:40 EST