Hi everyone,
Now that sfront 0.76:
http://www.cs.berkeley.edu/~lazzaro/sa/index.html
is out, it should be easier to continue testing the
PortAudio Windows functionality, to get a sense of how
practical it is to use sfront for real-time work under
Windows with these two APIs. Three of the examples:
examples/rtime/aatest ## streams a long MIDI file to sound output
examples/rtime/linain ## reverb unit using audio in and audio out
examples/rtime/linmidi ## real-time synth using -cin win32 for MIDI in
include a README file, that describes how to try to make
these work under Windows. A few items that didn't make it into
these READMEs:
----[1] If the default output or default input isn't the hardware port you want, set environment variables:
set PA_RECOMMENDED_OUTPUT_DEVICE=1 set PA_RECOMMENDED_INPUT_DEVICE=1
to cycle through the devices. Device 0 is the default device, higher numbers are alternative devices.
[2] You may find it useful to set the -latency flag as you experiment to find the underlying latency of the system, see
http://www.cs.berkeley.edu/~lazzaro/sa/sfman/user/cmdline/index.html#latency
for details on this flag.
-----
Let me know how well these examples work (perhaps via the mailing list, so everyone can get a sense of the current status) -- if it seems like Windows support in sfront 0.76 is actually usable and useful, then the route to getting sfront networking up and running is not that large, it would require:
-- Getting a version of the control driver ascii.c working -- Porting the networking code in sfront/src/lib/nsys/*.c to work under Windows. Since this known to work under many UNIX platforms now (Linux, Mac OS X, Solaris, HPUX), it might port to Windows quickly.
I'd suggest trying to compile the sfront/examples/nmp_null example, then taking the sa.c file and manually editing it for the obvious changes (changing include file names, etc). It might be easier to delete the -cin ascii driver at first, since that code (which turns the ASCII keyboard into a mono MIDI controller) is based on termios and would need a rewrite ... --jl
------------------------------------------------------------------------- John Lazzaro -- Research Specialist -- CS Division -- EECS -- UC Berkeley lazzaro [at] cs [dot] berkeley [dot] edu www.cs.berkeley.edu/~lazzaro -------------------------------------------------------------------------
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 11:46:43 EST