Thomas Huber  wrote:

> > depending on the audio signal, I render the files at 176400 Hz (4x),
> > 352800 Hz (8x) or 705600 Hz (16x). The only problem with this is that
> > some csound opcodes does not like high sample rates (especially
> > some filters)

> Why does the sampling rate have an effect on the function of an
> opcode ? I mean, the only thing that changes is the ratio of the
> processed frequencies to the sampling rate.

Some (buggy) opcodes may be tuned for one sampling rate (e.g. 44100Hz).
Also there may be precision problems, especially with filters; an example:

/* ftable 1 contains a sine wave */

afoo buzz sr/(10*3.14159265), 440, sr/(2*440), 1, 0
abar tone afoo, 10

This will render a 440Hz sawtooth wave at sr = 44100Hz, but after setting sr to
441000 Hz it will output silence

> > > 0 8192-byte soundblks ....
> > 
> > Try to use headerless (raw) file format. WAV needs seeking which is not supported
> > by pipes/fifos.
>
> I just tried. Same effect:
>
> soundfile write returned bytecount of -1, not 8192
> (disk may be full...
> closing the file ...)

hmmm... try this:

mkfifo foobar ; ( csound -d -m7 -h -s -H1 -o foobar blah.orc blah.sco & \
sox -t raw -s -w -c 1 -r 44100 foobar -t wav foobar.wav ) ; rm -f foobar

(in blah.orc sr = 44100 and nchnls = 1)

This is not resampling, but should work as an example of pipes/fifos.