This is designed to run a single patch, reading audio in from a file and saving the audio output to a file.
Usage:
cd simulator
cmake --fresh --preset headless
cmake --build build
build/simulator -p../patches/default/Djembe4Verb.yml
The -p option specifies the patch file to load.
You can provide it a stereo .wav file with the --in option. This will be sent
to In 1 and In 2. If you don't provide a file, then it will send silence to
those channels.
The output from Out 1 and Out 2 is saved to a .wav file, you can specify it
with --out or else a default name will be used.
You can specify the number of samples to process with -n, for example -n 48000 will process 1 second of sound at 48kHz.
Purpose: This project is useful for testing the audio stream while developing, for example, making sure that optimizations do not effect the audio outputs. It also displays the processing time as a percentage of sample time at 48kHz, which is useful for seeing if optimizations are having an effect (on the host computer, at least).
Finally, we've been using this already to compare the MM engine performance on various processors.
Future:
A future improvement would be to tie it into the OS's audio drivers, so it could stream audio. The driver adaptor should be done such that it can tie into the hardware MM's audio stream easily (so we can run headless on the MM).
With streaming audio working, it could then respond to MIDI events on the host computer, so you could use it as a full headless synth.