MIDI and SFX sequencer, designed for compatibility with MusyX audio groups
Go to file
Jack Andersen 695fc10b8f Setup ID prompt for amuserender 2016-07-06 18:19:40 -10:00
AudioUnit Work on MIDI-to-SNG conversion 2016-06-21 18:18:28 -10:00
VST Initial amuseconv implementation with SNG extraction 2016-06-20 20:40:13 -10:00
driver Setup ID prompt for amuserender 2016-07-06 18:19:40 -10:00
include/amuse Refactored audio supply dispatch across two passes 2016-07-04 15:08:00 -10:00
lib Refactored audio supply dispatch across two passes 2016-07-04 15:08:00 -10:00
.gitignore Initial Amuse AudioUnit frontend architecture 2016-05-24 16:00:22 -10:00
CMakeLists.txt Add amuserender executable 2016-07-06 11:30:46 -10:00
Doxyfile Add Doxyfile 2016-05-18 19:55:10 -10:00
LICENSE Add LICENSE 2016-05-02 11:54:23 -10:00
README.md Fix HQ reverb, better submix/sequencer allocation 2016-05-30 19:15:20 -10:00

README.md

Amuse

Amuse is a real-time MIDI and SFX sequencer, with basic effects, 3D positional audio and surround-output capabilities.

The project is designed for compatibility with Audio Groups and Song data found in PC/N64/GCN/GBA games using the MusyX audio engine; providing an alternate runtime library to use for sequencing these games' audio libraries.

Command-Line Player

Download

A simple command-line program for loading and playing AudioGroups out of game archives or raw (.proj,.pool,.sdir,.samp) files is provided.

[jacko@ghor ~]$ amuseplay test.proj
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
░░░   ████ ████  ┃  ████ ████ ████   ┃   ████ ████  ░░░
░░░   ████ ████  ┃  ████ ████ ████   ┃   ████ ████  ░░░
░░░   ▌W▐█ ▌E▐█  ┃  ▌T▐█ ▌Y▐█ ▌U▐█   ┃   ▌O▐█ ▌P▐█  ░░░
░░░    │    │    ┃    │    │    │    ┃    │    │    ░░░
░░░ A  │ S  │ D  ┃ F  │ G  │ H  │ J  ┃ K  │ L  │ ;  ░░░
░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
<left/right>: cycle MIDI setup, <up/down>: volume, <space>: PANIC
<tab>: sustain pedal, <window-Y>: pitch wheel, <window-X>: mod wheel
<Z/X>: octave, <C/V>: velocity, <B/N>: channel, <,/.>: program, <Q>: quit
  0 Setup 0, Chan 0, Prog 0, Octave: 4, Vel: 64, VOL: 80%

The command-line program requires a windowing environment and will open a small 100x100 window alongside your terminal/cmd. This window must be frontmost, since it listens to full keyboard events through it.

If a MIDI keyboard is connected and recognized by your OS before amuseplay is launched, you may directly control the sequencer using physical input.

On OS X and Linux, amuseplay will advertise a virtual MIDI-IN port for other audio applications to route their MIDI messages to. This enables tracker, drum machine, and DAW applications to produce sampled audio using Amuse directly.

Library

The Amuse API exposes full interactivity between a client application (game engine) and the sequencer engine. Unlike the interrupt-driven nature of the original console implementations (where the audio chip 'requests' more audio as needed), Amuse is entirely synchronous. This means the client must periodically pump the audio engine (typically once per video frame) to keep the OS' audio system fed.

The client must provide the implementation for allocating and mixing audio voices, since this may drastically differ from target to target. amuse::IBackendVoiceAllocator is the pure-virtual interface to implement for this. Alternatively, if Boo is present in the CMake project tree, Amuse will be compiled with a backend supporting multiple popular low-level audio APIs. Windows, OS X, and Linux all have excellent support this way.

Here's an example usage:

#include <amuse/amuse.hpp>
#include "MyVoiceAllocator.hpp"
#include "MyAudioGroupLoader.hpp"

int main(int argc, char* argv[])
{
    /* Up to the client to implement voice allocation and mixing */
    std::unique_ptr<amuse::IBackendVoiceAllocator> voxAlloc = MakeMyVoiceAllocator();

    /* Application just needs one per audio output (not per channel) */
    amuse::Engine snd(*voxAlloc);

    /* An 'AudioGroup' is an atomically-loadable unit within Amuse. 
     * A client-assigned integer serves as the handle to the group once loaded
     */
    amuse::IntrusiveAudioGroupData data = LoadMyAudioGroup();
    snd.addAudioGroup(data);

    /* Starting a SoundMacro playing is accomplished like so: */
    int sfxId = 0x1337;
    float vol = 1.0f;
    float pan = 0.0f;
    std::shared_ptr<amuse::Voice> voice = snd.fxStart(sfxId, vol, pan);

    /* Play for ~5 sec */
    int passedFrames = 0;
    while (passedFrames < 300)
    {
        snd.pumpEngine();
        ++passedFrames;
        WaitForVSync();
    }

    /* Stopping a SoundMacro is accomplished by sending a
     * MIDI-style 'KeyOff' message for the voice
     */
    voice->keyOff();
    
    /* Play for 2 more seconds to allow the macro to gracefully fade-out */
    passedFrames = 0;
    while (passedFrames < 120)
    {
        snd.pumpEngine();
        ++passedFrames;
        WaitForVSync();
    }

    /* Clean up and exit */
    return 0;
}