Csound can also render audio straight to a sound file stored on your hard drive instead of as live audio sent to the audio hardware. This gives you the possibility to hear the results of very complex processes which your computer can't produce in realtime.
Csound can render to formats like wav, aiff or ogg (and other less popular ones), but not mp3 due to its patent and licencing problems.
Save the following code as Render.csd:
EXAMPLE 02E01.csd
<CsoundSynthesizer>
<CsOptions>
-o Render.wav
</CsOptions>
<CsInstruments>
;Example by Alex Hofmann
instr 1
aSin oscils 0dbfs/4, 440, 0
out aSin
endin
</CsInstruments>
<CsScore>
i 1 0 1
e
</CsScore>
</CsoundSynthesizer>
Open the Terminal / Prompt / Console and type:
csound /path/to/Render.csd
Now, because you changed the -o flag in the <CsOptions> from "-o dac" to "-o filename", the audio output is no longer written in realtime to your audio device, but instead to a file. The file will be rendered to the default directory (usually the user home directory). This file can be opened and played in any audio player or editor, e.g. Audacity. (By default, csound is a non-realtime program. So if no command line options are given, it will always render the csd to a file called test.wav, and you will hear nothing in realtime.)
The -o flag can also be used to write the output file to a certain directory. Something like this for Windows ...
<CsOptions> -o c:/music/samples/Render.wav </CsOptions>
... and this for Linux or Mac OSX:
<CsOptions> -o /Users/JSB/organ/tatata.wav </CsOptions>
The internal rendering of audio data in Csound is done with 32-bit floating point numbers (or even with 64-bit numbers for the "double" version). Depending on your needs, you should decide the precision of your rendered output file:
For making sure that the header of your soundfile will be written correctly, you should use the -W flag for a WAV file, or the -A flag for a AIFF file. So these options will render the file "Wow.wav" as WAV file with 24-bit accuracy:
<CsOptions> -o Wow.wav -W -3 </CsOptions>
Sometimes you may want to simultaneously have realtime output and file rendering to disk, like recording your live performance. This can be achieved by using the fout opcode. You just have to specify your output file name. File type and format are given by a number, for instance 18 specifies "wav 24 bit" (see the manual page for more information). The following example creates a random frequency and panning movement of a sine wave, and writes it to the file "live_record.wav" (in the same directory as your .csd file):
EXAMPLE 02E02.csd
<CsoundSynthesizer>
<CsOptions>
-odac
</CsOptions>
<CsInstruments>
;Example by Joachim Heintz
sr = 44100
ksmps = 32
nchnls = 2
0dbfs = 1
seed 0 ;each time different seed for random
giSine ftgen 0, 0, 2^10, 10, 1 ;a sine wave
instr 1
kFreq randomi 400, 800, 1 ;random frequency
aSig poscil .2, kFreq, giSine ;sine with this frequency
kPan randomi 0, 1, 1 ;random panning
aL, aR pan2 aSig, kPan ;stereo output signal
outs aL, aR ;live output
fout "live_record.wav", 18, aL, aR ;write to soundfile
endin
</CsInstruments>
<CsScore>
i 1 0 10
e
</CsScore>
</CsoundSynthesizer>
All the options which are described in this chapter can be handled very easily in QuteCsound: