All the source code examples in this document are in the public domain; you can use them as you please. This document is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-sa/3.0/ . The FluidSynth library is distributed under the GNU Lesser General Public License. A copy of the GNU Lesser General Public License is contained in the FluidSynth package; if not, visit http://www.gnu.org/licenses/old-licenses/lgpl-2.1.txt or write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
<a href="http://www.fluidsynth.org">FluidSynth</a> is a software synthesizer based on the <a href="http://en.wikipedia.org/wiki/SoundFont">SoundFont 2</a> specifications. The synthesizer is available as a shared object that can easily be reused in any application that wants to use wave-table synthesis. This document explains the basic usage of FluidSynth. Some of the more advanced features are not yet discussed but will be added in future versions.
SoundFont(R) is a registered trademark of E-mu Systems, Inc.
\section Introduction
What is FluidSynth?
- FluidSynth is a software synthesizer based on the SoundFont 2 specifications. The synthesizer is available as a shared object (a concept also named Dynamic Linking Library, or DLL) that can be easily reused in any application for wave-table synthesis. This document explains the basic usage of FluidSynth.
- FluidSynth provides a Command Line Interface program ready to be used from the console terminal, offering most of the library functionalities to end users, among them the ability of render and play Standard MIDI Files, receive real-time MIDI events from external hardware ports and other applications, perform advanced routing of such events, enabling at the same time a local shell as well as a remote server commands interface.
- FluidSynth is an API (Application Programming Interface) relieving programmers from a lot of details of reading SoundFont and MIDI events and files, and sending the digital audio output to a Sound Card. These tasks can be accomplished using a small set of functions. This document explains most of the API functions and gives short examples about them.
- FluidSynth uses instrument samples contained in standard SF2 (SoundFont 2) files, having a file structure based on the RIFF format. The specification can be obtained here: http://connect.creativelabs.com/developer/SoundFont/Forms/AllItems.aspx but most users don't need to know any details of the format.
- FluidSynth can easily be embedded in an application. It has a main header file, fluidsynth.h, and one dynamically linkable library. FluidSynth runs on Linux, Mac OS X, and the Windows platforms, and support for OS/2 and OpenSolaris is experimental. It has audio and midi drivers for all mentioned platforms but you can use it with your own drivers if your application already handles MIDI and audio input/output. This document explains the basic usage of FluidSynth and provides examples that you can reuse.
- the following getters have been added: fluid_voice_is_on(), fluid_voice_is_sustained(), fluid_voice_is_sostenuto(), fluid_voice_get_channel(), fluid_voice_get_key(), fluid_voice_get_actual_key(), fluid_voice_get_velocity(), fluid_voice_get_actual_velocity(), fluid_player_get_current_tick(), fluid_player_get_total_ticks(), fluid_player_get_bpm(), fluid_player_get_midi_tempo()
Changes in FluidSynth 1.1.4 concerning developers:
- You can now play MIDI files that reside in memory (instead of specifying a filename). See \ref MIDIPlayerMem for an example.
- A hook can be inserted for MIDI file playback, at playback time. This is done through the new fluid_player_set_playback_callback API function. You can use this to both inspect and modify MIDI events as they are being played (or add a MIDI router), just as you can for MIDI input drivers.
- Channel 10 used to be the one and only drum channel, this can now be changed using the fluid_synth_set_channel_type.
- fluid_synth_all_sounds_off and fluid_synth_all_notes_off are new public API functions. You can use them to turn notes off (i e releasing all keys, voices advance to release phase) or sounds off (more like pressing the mute button), for one channel or all channels.
Before you can use the synthesizer, you have to create a settings object. The settings objects is used by many components of the FluidSynth library. It gives a unified API to set the parameters of the audio drivers, the midi drivers, the synthesizer, and so forth. A number of default settings are defined by the current implementation.
All settings have a name that follows the "dotted-name" notation. For example, "synth.polyphony" refers to the number of voices (polyphony) preallocated by the synthesizer. The settings also have a type. There are currently three types: strings, numbers (double floats), and integers. You can change the values of a setting using the fluid_settings_setstr(), fluid_settings_setnum(), and fluid_settings_setint() functions. For example:
The API contains the functions to query the type, the current value, the default value, the range and the "hints" of a setting. The range is the minimum and maximum value of the setting. The hints gives additional information about a setting. For example, whether a string represents a filename. Or whether a number should be interpreted on on a logarithmic scale. Check the settings.h API documentation for a description of all functions.
<br /><br /><strong>Deprecated:</strong><br />As of 1.1.7 this option is deprecated. This option enforces thread safety for rvoice_mixer, which causes rvoice_events to be queued internally. The current implementation relies on the fact that this option is set to TRUE to correctly render any amount of requested audio. Also calling fluid_synth_write_* in parallel is not considered to be a use-case. It would cause undefined audio output, as it would be unpredictable for the user which rvoice_events specifically would be dispatched to which fluid_synth_write_* call.
\section CreatingAudioDriver Creating the Audio Driver
The synthesizer itself does not write any audio to the audio output. This allows application developers to manage the audio output themselves if they wish. The next section describes the use of the synthesizer without an audio driver in more detail.
Creating the audio driver is straightforward: set the appropriate settings and create the driver object. Because the FluidSynth has support for several audio systems, you may want to change which one you want to use. The list below shows the audio systems that are currently supported. It displays the name, as used by the fluidsynth library, and a description.
- jack: JACK Audio Connection Kit (Linux, Mac OS X, Windows)
- alsa: Advanced Linux Sound Architecture (Linux)
- oss: Open Sound System (Linux, Unix)
- pulseaudio: PulseAudio (Linux, Mac OS X, Windows)
The default audio driver depends on the settings with which FluidSynth was compiled. You can get the default driver with fluid_settings_getstr_default(). To get the list of available drivers use the fluid_settings_foreach_option() function. Finally, you can set the driver with fluid_settings_setstr(). In most cases, the default driver should work out of the box.
Additional options that define the audio quality and latency are "audio.sample-format", "audio.period-size", and "audio.periods". The details are described later.
You create the audio driver with the new_fluid_audio_driver() function. This function takes the settings and synthesizer object as arguments. For example:
As soon as the audio driver is created, it will start playing. The audio driver creates a separate thread that uses the synthesizer object to generate the audio.
There are a number of general audio driver settings. The audio.driver settings define the audio subsystem that will be used. The audio.periods and audio.period-size settings define the latency and robustness against scheduling delays. There are additional settings for the audio subsystems used which are documented in another table.
<td>Jack server to connect to. Defaults to an empty string, which uses
default Jack server.</td>
</tr>
<tr>
<td>audio.oss.device</td>
<td>Type</td>
<td>string</td>
</tr>
<tr>
<td></td>
<td>Default</td>
<td>/dev/dsp</td>
</tr>
<tr>
<td></td>
<td>Description</td>
<td>Device to use for OSS audio output.</td>
</tr>
<tr>
<td>audio.portaudio.device</td>
<td>Type</td>
<td>string</td>
</tr>
<tr>
<td></td>
<td>Default</td>
<td>PortAudio Default</td>
</tr>
<tr>
<td></td>
<td>Description</td>
<td>Device to use for PortAudio driver output. Note that 'PortAudio Default'
is a special value which outputs to the default PortAudio device.</td>
</tr>
<tr>
<td>audio.pulseaudio.device</td>
<td>Type</td>
<td>string</td>
</tr>
<tr>
<td></td>
<td>Default</td>
<td>"default"</td>
</tr>
<tr>
<td></td>
<td>Description</td>
<td>Device to use for PulseAudio driver output</td>
</tr>
<tr>
<td>audio.pulseaudio.server</td>
<td>Type</td>
<td>string</td>
</tr>
<tr>
<td></td>
<td>Default</td>
<td>"default"</td>
</tr>
<tr>
<td></td>
<td>Description</td>
<td>Server to use for PulseAudio driver output</td>
</tr>
</table>
\section UsingSynth Using the synthesizer without an audio driver
It is possible to use the synthesizer object without creating an audio driver. This is desirable if the application using FluidSynth manages the audio output itself. The synthesizer has several API functions that can be used to obtain the audio output:
fluid_synth_write_s16() fills two buffers (left and right channel) with samples coded as signed 16 bits (the endian-ness is machine dependent). fluid_synth_write_float() fills a left and right audio buffer with 32 bits floating point samples. For multi channel audio output, the function fluid_synth_nwrite_float() has to be used.
The function fluid_synth_process() is still experimental and its use is therefore not recommended but it will probably become the generic interface in future versions.
\section LoadingSoundfonts Loading and managing SoundFonts
Before any sound can be produced, the synthesizer needs a SoundFont.
SoundFonts are loaded with the fluid_synth_sfload() function. The function takes the path to a SoundFont file and a boolean to indicate whether the presets of the MIDI channels should be updated after the SoundFont is loaded. When the boolean value is TRUE, all MIDI channel bank and program numbers will be refreshed, which may cause new instruments to be selected from the newly loaded SoundFont.
The synthesizer can load any number of SoundFonts. The loaded SoundFonts are treated as a stack, where each new loaded SoundFont is placed at the top of the stack. When selecting presets by bank and program numbers, SoundFonts are searched beginning at the top of the stack. In the case where there are presets in different SoundFonts with identical bank and program numbers, the preset from the most recently loaded SoundFont is used. The fluid_synth_program_select() can be used for unambiguously selecting a preset or bank offsets could be applied to each SoundFont with fluid_synth_set_bank_offset(), to try and ensure that each preset has unique bank and program numbers.
The fluid_synth_sfload() function returns the unique identifier of the loaded SoundFont, or -1 in case of an error. This identifier is used in subsequent management functions: fluid_synth_sfunload() removes the SoundFont, fluid_synth_sfreload() reloads the SoundFont. When a SoundFont is reloaded, it retains it's ID and position on the SoundFont stack.
Additional API functions are provided to get the number of loaded SoundFonts and to get a pointer to the SoundFont.
Once the synthesizer is up and running and a SoundFont is loaded, most people will want to do something useful with it. Make noise, for example. MIDI messages can be sent using the fluid_synth_noteon(), fluid_synth_noteoff(), fluid_synth_cc(), fluid_synth_pitch_bend(), fluid_synth_pitch_wheel_sens(), and fluid_synth_program_change() functions. For convenience, there's also a fluid_synth_bank_select() function (the bank select message is normally sent using a control change message).
FluidSynth can process real-time MIDI events received from hardware MIDI ports or other applications. To do so, the client must create a MIDI input driver. It is a very similar process to the creation of the audio driver: you initialize some properties in a settings instance and call the new_fluid_midi_driver() function providing a callback function that will be invoked when a MIDI event is received. The following MIDI drivers are currently supported:
- jack: JACK Audio Connection Kit MIDI driver (Linux, Mac OS X)
- oss: Open Sound System raw MIDI (Linux, Unix)
- alsa_raw: ALSA raw MIDI interface (Linux)
- alsa_seq: ALSA sequencer MIDI interface (Linux)
- winmidi: Microsoft Windows MM System (Windows)
- midishare: MIDI Share (Linux, Mac OS X)
- coremidi: Apple CoreMIDI (Mac OS X)
\code
#include <fluidsynth.h>
int handle_midi_event(void* data, fluid_midi_event_t* event)
FluidSynth can be used to play MIDI files, using the MIDI File Player interface. It follows a high level implementation, though its implementation is currently incomplete. After initializing the synthesizer, create the player passing the synth instance to new_fluid_player(). Then, you can add some SMF file names to the player using fluid_player_add(), and finally call fluid_player_play() to start the playback. You can check if the player has finished by calling fluid_player_get_status(), or wait for the player to terminate using fluid_player_join().
<td>If true, reset the synth before starting a new MIDI song, so the state of a previous song can't affect the new song. Turn it off for seamless looping of a song.</td>
\section MIDIPlayerMem Playing a MIDI file from memory
FluidSynth can be also play MIDI files directly from a buffer in memory. If you need to play a file from a stream (such as stdin, a network, or a high-level file interface), you can load the entire file into a buffer first, and then use this approach. Use the same technique as above, but rather than calling fluid_player_add(), load it into memory and call fluid_player_add_mem() instead. Once you have passed a buffer to fluid_player_add_mem(), it is copied, so you may use it again or free it immediately (it is your responsibility to free it if you allocated it).
The MIDI router is one more processing layer directly behind the MIDI input. It processes incoming MIDI events and generates control events for the synth. It can be used to filter or modify events prior to sending them to the synthesizer. When created, the MIDI router is transparent and simply passes all MIDI events. Router "rules" must be added to actually make use of its capabilities.
FluidSynth's sequencer can be used to play MIDI events in a more flexible way than using the MIDI file player, which expects the events to be stored as Standard MIDI Files. Using the sequencer, you can provide the events one by one, with an optional timestamp for scheduling.
The client program should first initialize the sequencer instance using the function new_fluid_sequencer2(). There is a complementary function delete_fluid_sequencer() to delete it. After creating the sequencer instance, the destinations can be registered using fluid_sequencer_register_fluidsynth() for the synthesizer destination, and optionally using fluid_sequencer_register_client() for the client destination providing a suitable callback function. It can be unregistered using fluid_sequencer_unregister_client(). After the initialization, events can be sent with fluid_sequencer_send_now() and scheduled to the future with fluid_sequencer_send_at(). The registration functions return identifiers, that can be used as destinations of an event using fluid_event_set_dest().
The function fluid_sequencer_get_tick() returns the current playing position. A program may choose a new timescale in milliseconds using fluid_sequencer_set_time_scale().
The following example uses the fluidsynth sequencer to implement a sort of music box. FluidSynth internal clock is used to schedule repetitive sequences of notes. The next sequence is scheduled on advance before the end of the current one, using a timer event that triggers a callback function. The scheduling times are always absolute values, to avoid slippage.
\code
#include "fluidsynth.h"
fluid_synth_t* synth;
fluid_audio_driver_t* adriver;
fluid_sequencer_t* sequencer;
short synthSeqID, mySeqID;
unsigned int now;
unsigned int seqduration;
// prototype
void seq_callback(unsigned int time, fluid_event_t* event, fluid_sequencer_t* seq, void* data);
The shell interface allows you to send simple textual commands to the synthesizer, to parse a command file, or to read commands from the stdin or other input streams. To find the list of currently supported commands, please check the fluid_cmd.c file or type "help" in the fluidsynth command line shell.
