sdl2.sdlmixer - Python bindings for SDL_mixer¶
The sdl2.sdlmixer
module provides bindings for wrapper the SDL_mixer
audio playback library. It supports loading, playing, and mixing audio from a
wide range of popular formats (e.g. WAV, FLAC, MP3, OGG).
The SDL_mixer library is designed around the concept of virtual channels, each of which can be set to different volumes, have different left/right panning, and play different effects and sounds simultaneously (like a mixing board in a recording studio). In addition to the main mixing channels, SDL_mixer also provides a single channel for music playback, which is less flexible than the other channels but supports a wider range of playback formats (including MIDI) and requires less memory for playback.
Music playback is performed using Mix_Music
objects and their
corresponding functions. All other audio playback is performed using
Mix_Chunk
objects.
Table Of Contents
Note
The PySDL2 documentation for this module is currently a work-in-progress, and not all functions are documented. The official C documentation (linked below) can be used in the meantime to fill in any gaps.
The official documentation for the SDL_mixer library can be found here. Note that not all functions may be documented, as the official doucmentation is currently several releases out-of-date.
Main data types¶
-
class
sdl2.sdlmixer.
Mix_Chunk
[source]¶ A loaded audio clip to use for playback with the mixer API.
Chunk objects are created by the
Mix_LoadWAV()
andMix_QuickLoad
functions, and should be freed usingMix_FreeChunk()
when no longer needed.-
allocated
¶ Whether the associated audio buffer will be freed when the chunk is freed.
1
if the buffer is owned by the chunk, or0
if it was allocated by a different function.Type: int
-
-
class
sdl2.sdlmixer.
Mix_Music
[source]¶ The opaque data type representing a loaded music file.
Music objects are created by the
Mix_LoadMUS()
family of functions and should be freed usingMix_FreeMusic()
when no longer needed.
Initialization and library information functions¶
-
sdl2.sdlmixer.
Mix_Init
(flags)[source]¶ Initializes the SDL2_mixer library.
Calling this function enables support for the FLAC, MOD, MP3, and/or Ogg Vorbis audio formats as requested by the init flags. All other audio file formats can be loaded or used regardless of whether this has been called.
The following init flags are supported:
Format Init flag FLAC MIX_INIT_FLAC
MOD MIX_INIT_MID
MP3 MIX_INIT_MP3
Ogg Vorbis MIX_INIT_OGG
This can be called multiple times to enable support for these formats separately, or can initialize multiple formats at once by passing a set of flags as a bitwise OR. You can also call this function with 0 as a flag to check which audio decoding libraries have already been loaded, or to test whether a given decoder is available on the current system:
# Initialize FLAC and MP3 support separately for flag in [MIX_INIT_FLAC, MIX_INIT_MP3]: Mix_Init(flag) err = Mix_GetError() # check for any errors loading library if len(err): print(err) # Initialize FLAC and MP3 support at the same time flags = MIX_INIT_FLAC | MIX_INIT_MP3 Mix_Init(flags) if Mix_Init(0) != flags: # verify both libraries loaded properly print(Mix_GetError())
Note
This function is not guaranteed to set an error string on failure, so the return value should be used for error checking instead of just
Mix_GetError()
.Parameters: flags (int) – A bitwise OR’d set of the flags of the audio formats to load support for. Returns: A bitmask of all the currently initialized audio decoders. Return type: int
-
sdl2.sdlmixer.
Mix_Quit
()[source]¶ De-initializes the SDL2_mixer library.
Calling this function disables FLAC, MOD, MP3, and Ogg support and frees all associated memory. Once this has been called, you can re-initialize support for those audio decoders using
Mix_Init()
and the corresponding init flags.You only need to call this function once, no matter how many times
Mix_Init()
was called.
-
sdl2.sdlmixer.
Mix_OpenAudio
(frequency, format, channels, chunksize)[source]¶ Opens the default audio output device for use with the mixer API.
This function opens the default audio device with a given output channel count, audio sample format, sample rate, and audio buffer size, and initializes the mixer with 8 virtual channels:
# Initialize a 44.1 kHz 16-bit stereo mixer with a 1024-byte buffer size ret = Mix_OpenAudio(44100, sdl2.AUDIO_S16SYS, 2, 1024) if ret < 0: err = Mix_GetError().decode("utf8") raise RuntimeError("Error initializing the mixer: {0}".format(err))
A sample rate of 44100 Hz (CD quality) or 48000 Hz is recommended for any remotely modern computer. The chunk size must be a power of 8 (e.g. 512, 1024, 2048), and specifies how much audio data to send to the output device at a time. Lower values will have lower latencies, but may introduce skips in the audio (2048 is a safe default). Supported channel counts are 1 (mono), 2 (stereo), 4 (quad), and 6 (5.1 surround).
For a full list of supported audio format flags, see the list of format values at the following link: https://wiki.libsdl.org/SDL_AudioSpec
Note
This should not be called if
SDL_OpenAudio()
has already opened the default audio device (and vice-versa).Parameters: - frequency (int) – The output sampling rate per channel (in Hz).
- format (int) – A constant indicating the audio sample format to use
with the device (e.g.
sdl2.AUDIO_S16
). - channels (int) – The number of output channels to use for the device (e.g. 2 for stereo, 1 for mono).
- chunksize (int) – The size of the audio buffer (in bytes).
Returns: 0 on success, or -1 on error.
Return type:
-
sdl2.sdlmixer.
Mix_OpenAudioDevice
(frequency, format, channels, chunksize, device, allowed_changes)[source]¶ Opens a specific audio output device for use with the mixer API.
A specific audio device can be opened by name using the output of the
SDL_GetAudioDeviceName()
function. Alternatively, passingNone
as the device name will make open the most reasonable default audio device.The
allowed_changes
flag specifies which output properties (channel count, frequency, sample format) are allowed to be be automatically changed if not supported by the chosen audio device. See the documentation forSDL_OpenAudioDevice()
for the list of supported flags and more info.See
Mix_OpenAudio()
for more usage details.Note
Once an output device has been opened with this function, it should not be opened again with
SDL_OpenAudioDevice()
(or vice-versa).Parameters: - frequency (int) – The output sampling rate per channel (in Hz).
- format (int) – A constant indicating the audio sample format to use
with the device (e.g.
sdl2.AUDIO_S16
). - channels (int) – The number of output channels to use for the device (e.g. 2 for stereo, 1 for mono).
- chunksize (int) –
- device (bytes) – A UTF-8 encoded bytestring of the name of the audio
output device to open (or
None
for the default). - allowed_changes (int) – A bitmask of flags indicating the output properties allowed to be automatically changed to support the output device. If 0, no changes are allowed.
Returns: 0 on success, or -1 on error.
Return type:
-
sdl2.sdlmixer.
Mix_CloseAudio
()[source]¶ Shuts down and de-initializes the mixer API.
Calling this function stops all audio playback and closes the current mixer device. Once called, the mixer API should not be used until re-initialized with
Mix_OpenAudioDevice()
.Note
If
Mix_OpenAudioDevice()
has been called multiple times, this must be called an equal number of times to actually de-initialize the API.
-
sdl2.sdlmixer.
Mix_GetError
()¶ Returns the most recently encountered SDL2 error message, if any.
This function is a simple wrapper around
SDL_GetError()
.Retuns: A UTF-8 encoded string describing the most recent SDL2 error. Return type: bytes
-
sdl2.sdlmixer.
Mix_SetError
(fmt)¶ Sets the most recent SDL2 error message to a given string.
This function is a simple wrapper around
SDL_SetError()
.Parameters: fmt (bytes) – A UTF-8 encoded string containing the error message to set. Retuns: Always returns -1
.Return type: int
-
sdl2.sdlmixer.
Mix_QuerySpec
(frequency, format, channels)[source]¶ Retrieves the actual audio format in use by the current mixer device.
This function returns the calculated info by reference, meaning that it needs to be called using pre-allocated ctypes variables:
from ctypes import c_int, byref freq, fmt, chans = c_int(0), c_int(0), c_int(0) ret = Mix_QuerySpec(byref(freq), byref(fmt), byref(chans)) if ret > 0: results = [x.value for x in (freq, fmt, chans)]
The obtained values may or may not match the parameters you passed to
Mix_OpenAudio()
, depending on the audio configurations supported by the output device.Parameters: Returns: The number of times the mixer device has been opened, or 0 on error.
Return type:
Audio loading and closing functions¶
Loading and closing audio clips¶
-
sdl2.sdlmixer.
Mix_LoadWAV
(file)[source]¶ Loads an audio clip from a file.
See
Mix_GetChunkDecoder()
for a list of supported file types.Note
Mix_OpenAudioDevice()
must be called before this function can be used in order to determine the correct output format for playback.Parameters: file (bytes) – A UTF8-encoded bytestring containing the path of the audio clip to load. Returns: A pointer to the chunk containing the loaded audio. Return type: POINTER( Mix_Chunk
)
-
sdl2.sdlmixer.
Mix_LoadWAV_RW
(src, freesrc)[source]¶ Loads an audio clip from an SDL2 file object.
See
Mix_GetChunkDecoder()
for a list of supported file types.Note
Mix_OpenAudioDevice()
must be called before this function can be used in order to determine the correct output format for playback.Parameters: - src (
SDL_RWops
) – A file object containing a valid audio clip. - freesrc (int) – If non-zero, the provided file object will be closed and freed automatically after being loaded.
Returns: A pointer to the chunk containing the loaded audio.
Return type: POINTER(
Mix_Chunk
)- src (
-
sdl2.sdlmixer.
Mix_QuickLoad_WAV
(mem)[source]¶ Loads a memory buffer as a WAV file.
Unlike
Mix_LoadWAV_RW()
, this function performs no audio format conversion or error checking, and assumes that the WAV in the buffer is already in the correct output format for the mixer. Unless high performance is a must,Mix_LoadWAV_RW()
is a more flexible and much safer option.Note
Mix_OpenAudioDevice()
must be called before this function can be used.Parameters: mem (POINTER( c_byte
)) – A pointer to a memory buffer containing a valid WAVE audio file.Returns: A pointer to the chunk containing the loaded audio. Return type: POINTER( Mix_Chunk
)
-
sdl2.sdlmixer.
Mix_QuickLoad_RAW
(mem, len)[source]¶ Loads a memory buffer as a raw audio clip.
This function performs no error checking and assumes that the data in the buffer is in the correct output format for the mixer.
This can be used for converting Numpy arrays or other Python data types into audio clips for use with the mixer API. For example, to generate a pure sine wave tone at a given frequency, you could use the following code:
import ctypes import numpy as np duration = 3 # Seconds of sound hz = 432 # Frequency of the generated tone dtype = np.int16 # Mixer output format is signed 16-bit int max_int = 32767 # The max/min value for a signed 16-bit int srate = 44100 # Sample rate for each channel is 44100 kHz # Generate a 3 sec. sine wave for a 2-channel, 44100 Hz, AUDIO_S16 mixer size = int((duration / 1000.0) * srate) arr = np.sin(np.pi * np.arange(size) / srate * hz) * max_int # Cast the array into ctypes format for use with mixer arr_bytes = arr.tostring() buflen = len(arr_bytes) c_buf = (ctypes.c_ubyte * buflen).from_buffer_copy(arr_bytes) # Convert the ctypes memory buffer into a mixer audio clip sine_chunk = Mix_QuickLoad_RAW( ctypes.cast(c_buf, ctypes.POINTER(ctypes.c_ubyte)), buflen )
Note
You must keep a reference to the created ctypes buffer as long as the resulting audio clip is in use. Otherwise, Python may automatically free the memory associated with the audio buffer, meaning that any subsequent attempts to play the clip will result in a segmentation fault.
Note
Mix_OpenAudioDevice()
must be called before this function can be used.Parameters: Returns: A pointer to the chunk containing the loaded audio.
Return type: POINTER(
Mix_Chunk
)
Loading and closing music¶
-
sdl2.sdlmixer.
Mix_LoadMUS
(file)[source]¶ Loads music from a file.
See
Mix_GetMusicDecoder()
for a list of supported file types.Note
Mix_OpenAudioDevice()
must be called before this function can be used in order to determine the correct output format for playback.Parameters: file (bytes) – A UTF8-encoded bytestring containing the path of the music file to load. Returns: A pointer to the object containing the loaded music. Return type: POINTER( Mix_Music
)
-
sdl2.sdlmixer.
Mix_LoadMUS_RW
(src, freesrc)[source]¶ Loads music from an SDL2 file object.
See
Mix_GetMusicDecoder()
for a list of supported file types.Note
Mix_OpenAudioDevice()
must be called before this function can be used in order to determine the correct output format for playback.Parameters: - src (
SDL_RWops
) – A file object containing a valid music format. - freesrc (int) – If non-zero, the provided file object will be closed and freed automatically after being loaded.
Returns: A pointer to the object containing the loaded music.
Return type: POINTER(
Mix_Music
)- src (
-
sdl2.sdlmixer.
Mix_LoadMUSType_RW
(src, type, freesrc)[source]¶ Loads music from an SDL2 file object with a specific decoder.
This function supports the following audio format constants:
Format Constant None (autodetect) MUS_NONE
External command MUS_CMD
WAVE format MUS_WAV
Amiga MOD format MUS_MOD
MIDI format MUS_MID
Ogg Vorbis MUS_OGG
MP3 format MUS_MP3
FLAC format MUS_FLAC
Opus format MUS_OPUS
Note
Mix_OpenAudioDevice()
must be called before this function can be used in order to determine the correct output format for playback.Parameters: Returns: A pointer to the object containing the loaded music.
Return type: POINTER(
Mix_Music
)
Decoder availability & info functions¶
-
sdl2.sdlmixer.
Mix_GetNumChunkDecoders
()[source]¶ Retrieves the number of available audio chunk decoders.
The returned value can differ between runs of a program due to changes in the availability of the shared libraries required for supporting different formats.
Returns: The number of available audio chunk decoders. Return type: int
-
sdl2.sdlmixer.
Mix_GetChunkDecoder
(index)[source]¶ Retrieves the name of a given audio chunk decoder.
The SDL_mixer library currently supports the following chunk decoders:
Decoder Name Format Type Notes b”FLAC” Free Lossless Audio Codec b”MOD” Amiga MOD format b”MP3” MP3 format b”OGG” Ogg Vorbis b”MID” MIDI format Not always available on Linux b”OPUS” Opus Interactive Audio Codec Added in SDL_mixer 2.0.4 b”WAVE” Waveform Audio File Format b”AIFF” Audio Interchange File Format b”VOC” Creative Voice file Use the
Mix_GetNumChunkDecoders()
function to get the number of available chunk decoders.Returns: The name of the given chunk decoder, or None
if the index is invalid.Return type: bytes
-
sdl2.sdlmixer.
Mix_HasChunkDecoder
(name)[source]¶ Checks whether a specific chunk decoder is available.
See
Mix_GetChunkDecoder()
for a list of valid decoder names.Parameters: name (bytes) – A bytestring of the name of the decoder to query. Returns: 1 if the decoder is present, or 0 if unavailable. Return type: int
-
sdl2.sdlmixer.
Mix_GetNumMusicDecoders
()[source]¶ Retrieves the number of available music decoders.
The returned value can differ between runs of a program due to changes in the availability of the shared libraries required for supporting different formats.
Returns: The number of available music decoders. Return type: int
-
sdl2.sdlmixer.
Mix_GetMusicDecoder
(index)[source]¶ Retrieves the name of a given music decoder.
The SDL_mixer library currently supports the following music decoders:
Decoder Name Format Type Notes b”FLAC” Free Lossless Audio Codec b”MODPLUG” Amiga MOD (via libmodplug) b”MOD” Amiga MOD (via libmikmod) b”MP3” MP3 format b”MPG123” MP3 format (duplicate) b”OGG” Ogg Vorbis b”MIDI” MIDI format Not always available on Linux b”NATIVEMIDI” MIDI format (OS native) Not available on Linux b”OPUS” Opus Interactive Audio Codec Added in SDL_mixer 2.0.4 b”CMD External music command Not available on Windows b”WAVE” Waveform Audio File Format Use the
Mix_GetNumMusicDecoders()
function to get the number of available chunk decoders.Returns: The name of the given music decoder, or None
if the index is invalid.Return type: bytes
-
sdl2.sdlmixer.
Mix_GetMusicType
(music)[source]¶ Gets the format of a given music object.
See
Mix_LoadMUSType_RW()
for a list of the possible type constants.Parameters: music ( Mix_Music
) – The music object for which the type will be retrieved.Returns: A constant indicating the format of the music object, or MUS_NONE
(0) if the format could not be identified.Return type: int
Channel functions¶
These functions affect regular mixer channels. Music is not affected by these functions.
Channel configuration¶
-
sdl2.sdlmixer.
Mix_AllocateChannels
(numchans)[source]¶ Sets the number of channels to use for the mixer API.
In this context, “channels” refers to the number of virtual channels used by the mixer API for playing multiple sounds simultaneously. It does not refer to the physical number of channels to use for the output device.
This can be called multiple times, even with sounds playing. If
numchans
is less than the current number of channels, the channels above the new number will be stopped, freed, and not mixed any longer. If any channels are deallocated, any callback set byMix_ChannelFinished()
will be called when each channel is halted to be freed.If this function is called with a negative number (e.g.
-1
), it will return the number of currently-allocated virtual mixer channels without changing anything. If called with 0, all mixer channels will be freed.This function has no effect on music playback.
Parameters: numchans (int) – The number of virtual channels to use for the mixer API, or a negative number to query the current allocated channel count.