naturau-base or ntrb, is the base (core) code of naturau, a full-scale dynamic audio mixer I'm working on; and a foundation for my future work in audio. naturau-base provides the foundation for playing and streaming audio from files (currently WAV or FLAC) or creating a source of your own, as well as a few utilities to aid development of other projects using naturau-base.
This is an example of a high-level audio playback with naturau-base:
#include "ntrb/alloc.h"
#include "ntrb/audeng_wrapper.h"
#include "ntrb/RuntimeCoreData.h"
#include <assert.h>
#include <stdbool.h>
int main(){
#ifdef NTRB_MEMDEBUG
assert(ntrb_memdebug_init_with_return_value() == ntrb_memdebug_OK);
#endif
ntrb_RuntimeCoreData rcd;
const size_t track_count = 2;
assert(ntrb_RuntimeCoreData_new(&rcd, track_count) == ntrb_RuntimeCoreData_OK);
const char* const aud_filename = "audio.wav";
ntrb_RuntimeCoreData_queue_audio(&rcd, aud_filename);
rcd.in_pause_state = false;
ntrb_run_audio_engine(&rcd);
ntrb_RuntimeCoreData_free(&rcd);
#ifdef NTRB_MEMDEBUG
assert(ntrb_memdebug_uninit(true) == ntrb_memdebug_OK);
#endif
return 0;
}Although, naturau-base is more designed towards being extended, modified or included in your projects. See naturau. There is truly no right or wrong way of using naturau-base. As long as you are able to make what you want, all the power to you :D
As mentioned above, no right or wrong way of using ntrb. If you need examples of extensions or modifications see naturau. Use ntrb's original code as a guideline for how things should be done, then extend or modify it to fit your code.
Here's an example of naturau_AudioTrack, essentially an ntrb_AudioBuffer with audio loudness multiplier and an audio effector struct:
typedef struct{
ntrb_AudioBuffer output_buffer;
float amplitude_multiplier;
naturau_AudioEffector audio_effector;
} naturau_AudioTrack;Modifications to audeng_wrapper and RuntimeCoreData are also present in naturau, since the original revolves around the ntrb_AudioBuffer struct.
This is a verbatim of the building documentation in the Doxygen documentation
Navigate to ./doc directory in ntrb directory, call doxygen ./docconf.
Anywhere you'd like. Just make sure you are able to build both of them.
You can try building PortAudio from scratch by following the guide from PortAudio, or in MSYS2 MinGW64, you can download the package.
Tip for people manually compiling with MSYS2 MinGW64: Remove the -DPA_*=0 arguments when creating CMake config. This ensures PortAudio will have multiple audio hosts for fallback.
Once that is completed, test the audio engine by compiling the examples in the PortAudio directory,
paex_sine.c or paex_saw.c is a good one.
To compile the an example in a single gcc call,
you need to provide the PortAudio include folder if needed with -I,
include the PortAudio lib file with -lportaudio and provide the directory of the lib file if needed
using the -L. My gcc call in the examples directory looks something like
gcc ./paex_sine.c -I../include -L../build -lportaudioIf you chose to install PortAudio as an MSYS package, the libPortAudio.dll file is in your environment. But when shipping your code, make sure to include it in where the compiler is called in your project directory as well.
If you chose to manually build one yourself, the .dll file is in the build directory of PortAudio
and must be present in the compiler's call directory.
You can either put the compiled .dll in your terminal's environment directly,
or put it in where you would call make or gcc in your project.
For your example compilation, you can either compile from the build directory instead or copy the .dll to the examples directory.
If the playback is correct, you're good on this one!
Follow the instructions on the github page of libFLAC. Download the release source code, not the platform specific release. We need to build it from scratch.
There are a few optimisations for building libFLAC in the CMakeLists.txt file in its directory.
If you don't have OGG libary from Xiph, switch WITH_OGG to OFF. Not doing so will cause an error.
Tip for MSYS MinGW64 users: if needed,
add -G "MSYS Makefiles" in cmake /path/to/flac-source to build with make.
Anything else should be left as is. But make sure BUILD_SHARED_LIBS is OFF. We are compiling it statically.
You can choose to make install libFLAC or not. It slightly changes how you incorporate it with ntrb,
but we got you covered for both :D
Once the libFLAC test passed, you are now clear to build ntrb.
ntrb itself doesn't have an executable, building ntrb means building the object files for other code to include.
First, rename or copy the "extern_path - template.make" to "extern_path.make".
In it, there are PORTAUDIO_INCLUDEDIR, PORTAUDIO_LIBDIR, FLAC_INCLUDEDIR, FLAC_LIBDIR.
You will need to fill in path to the include directory of PortAudio, the directory of the libPortAudio.a,
the include directory of libFLAC and the directory of libFLAC.a (in where-you-built-flac/src/libFLAC) respectively.
If you used the PortAudio package instead, the two probably don't have to be filled in.
And if you ran make install for libFLAC, you probably don't have to fill in the latter two.
To build ntrb, call make in the ntrb directory.
Aside from a few pointer warnings from the alloc module, no errors should occur,
and the ALL CLEAR for the unit test should be there.
Congratulations! You have built ntrb! You can try to build the example in the ./example directory.
Call make in there and you should have a working demonstration of ntrb code.
Or you can refer to the next page for including and building with ntrb in your project.