SDL3pp
A slim C++ wrapper for SDL3
|
SDL Audio Device instance IDs. More...
Public Member Functions | |
AudioDeviceBase (const AudioDeviceBase &devid, OptionalRef< const SDL_AudioSpec > spec) | |
Open a specific audio device. | |
constexpr auto | operator<=> (const AudioDeviceBase &other) const |
Comparison. | |
const char * | GetName () const |
Get the human-readable name of a specific audio device. | |
AudioSpec | GetFormat (int *sample_frames=nullptr) const |
Get the current audio format of a specific audio device. | |
OwnArray< int > | GetChannelMap () const |
Get the current channel map of an audio device. | |
bool | IsPhysical () const |
Determine if an audio device is physical (instead of logical). | |
bool | IsPlayback () const |
Determine if an audio device is a playback device (instead of recording). | |
void | Pause () |
Use this function to pause audio playback on a specified device. | |
void | Resume () |
Use this function to unpause audio playback on a specified device. | |
bool | Paused () const |
Use this function to query if an audio device is paused. | |
float | GetGain () const |
Get the gain of an audio device. | |
void | SetGain (float gain) |
Change the gain of an audio device. | |
void | BindAudioStreams (std::span< AudioStreamRef > streams) |
Bind a list of audio streams to an audio device. | |
void | BindAudioStream (AudioStreamBase &stream) |
Bind a single audio stream to an audio device. | |
void | SetPostmixCallback (AudioPostmixCB callback) |
Set a callback that fires when data is about to be fed to an audio device. | |
void | SetPostmixCallback (AudioPostmixCallback callback, void *userdata) |
Set a callback that fires when data is about to be fed to an audio device. | |
constexpr | Resource (T resource={}) |
Constructs the underlying resource. | |
constexpr | Resource (std::nullptr_t) |
Equivalent to default ctor. | |
constexpr | Resource (std::nullopt_t) |
Equivalent to default ctor. | |
Resource (const Resource &other)=delete | |
Resource (Resource &&other)=delete | |
![]() | |
constexpr | Resource (SDL_AudioDeviceID resource={}) |
Constructs the underlying resource. | |
constexpr | Resource (std::nullptr_t) |
Equivalent to default ctor. | |
constexpr | Resource (std::nullopt_t) |
Equivalent to default ctor. | |
Resource (const Resource &other)=delete | |
Resource (Resource &&other)=delete | |
Resource & | operator= (const Resource &other)=delete |
Resource & | operator= (Resource &&other)=delete |
constexpr | operator bool () const |
True if contains a valid resource. | |
constexpr bool | operator== (const Resource &other) const=default |
Comparison. | |
constexpr bool | operator== (std::nullopt_t) const |
Comparison. | |
constexpr bool | operator== (std::nullptr_t) const |
Comparison. | |
constexpr SDL_AudioDeviceID | get () const |
Return contained resource;. | |
constexpr SDL_AudioDeviceID | release (SDL_AudioDeviceID newResource={}) |
Return contained resource and empties or replace value. | |
constexpr const SDL_AudioDeviceID | operator-> () const |
Access to fields. | |
constexpr SDL_AudioDeviceID | operator-> () |
Access to fields. | |
Zero is used to signify an invalid/null device.
|
inline |
You can open both playback and recording devices through this function. Playback devices will take data from bound audio streams, mix it, and send it to the hardware. Recording devices will feed any bound audio streams with a copy of any incoming data.
An opened audio device starts out with no audio streams bound. To start audio playing, bind a stream and supply audio data to it. Unlike SDL2, there is no audio callback; you only bind audio streams and make sure they have data flowing into them (however, you can simulate SDL2's semantics fairly closely by using AudioStreamBase.AudioStreamBase instead of this function).
If you don't care about opening a specific device, pass a devid
of either AUDIO_DEVICE_DEFAULT_PLAYBACK
or AUDIO_DEVICE_DEFAULT_RECORDING
. In this case, SDL will try to pick the most reasonable default, and may also switch between physical devices seamlessly later, if the most reasonable default changes during the lifetime of this opened device (user changed the default in the OS's system preferences, the default got unplugged so the system jumped to a new default, the user plugged in headphones on a mobile device, etc). Unless you have a good reason to choose a specific device, this is probably what you want.
You may request a specific format for the audio device, but there is no promise the device will honor that request for several reasons. As such, it's only meant to be a hint as to what data your app will provide. Audio streams will accept data in whatever format you specify and manage conversion for you as appropriate. AudioDeviceBase.GetFormat can tell you the preferred format for the device before opening and the actual format the device is using after opening.
It's legal to open the same device ID more than once; each successful open will generate a new logical AudioDeviceBase that is managed separately from others on the same physical device. This allows libraries to open a device separately from the main app and bind its own streams without conflicting.
It is also legal to open a device ID returned by a previous call to this function; doing so just creates another logical device on the same physical device. This may be useful for making logical groupings of audio streams.
This function returns the opened device ID on success. This is a new, unique AudioDeviceBase that represents a logical device.
Some backends might offer arbitrary devices (for example, a networked audio protocol that can connect to an arbitrary server). For these, as a change from SDL2, you should open a default device ID and use an SDL hint to specify the target if you care, or otherwise let the backend figure out a reasonable default. Most backends don't offer anything like this, and often this would be an end user setting an environment variable for their custom need, and not something an application should specifically manage.
When done with an audio device, possibly at the end of the app's life, one should call AudioDeviceRef.reset() on the returned device id.
devid | the device instance id to open, or AUDIO_DEVICE_DEFAULT_PLAYBACK or AUDIO_DEVICE_DEFAULT_RECORDING for the most reasonable default device. |
spec | the requested device configuration. Can be nullptr to use reasonable defaults. |
Error | on failure. |
|
inline |
Channel maps are optional; most things do not need them, instead passing data in the order that SDL expects.
Audio devices usually have no remapping applied. This is represented by returning nullptr, and does not signify an error.
|
inline |
For an opened device, this will report the format the device is currently using. If the device isn't yet opened, this will report the device's preferred format (or a reasonable default if this can't be determined).
You may also specify AUDIO_DEVICE_DEFAULT_PLAYBACK or AUDIO_DEVICE_DEFAULT_RECORDING here, which is useful for getting a reasonable recommendation before opening the system-recommended default device.
You can also use this to request the current device buffer size. This is specified in sample frames and represents the amount of data SDL will feed to the physical hardware in each chunk. This can be converted to milliseconds of audio with the following equation:
ms = (int) ((((Sint64) frames) * 1000) / spec.freq);
Buffer size is only important if you need low-level control over the audio playback timing. Most apps do not need this.
sample_frames | pointer to store device buffer size, in sample frames. Can be nullptr. |
Error | on failure. |
|
inline |
The gain of a device is its volume; a larger gain means a louder output, with a gain of zero being silence.
Audio devices default to a gain of 1.0f (no change in output).
Physical devices may not have their gain changed, only logical devices, and this function will always return -1.0f when used on physical devices.
|
inline |
|
inline |
An AudioDeviceBase that represents physical hardware is a physical device; there is one for each piece of hardware that SDL can see. Logical devices are created by calling AudioDeviceBase.AudioDeviceBase or AudioStreamBase.AudioStreamBase, and while each is associated with a physical device, there can be any number of logical devices on one physical device.
For the most part, logical and physical IDs are interchangeable–if you try to open a logical device, SDL understands to assign that effort to the underlying physical device, etc. However, it might be useful to know if an arbitrary device ID is physical or logical. This function reports which.
This function may return either true or false for invalid device IDs.
|
inline |
This function may return either true or false for invalid device IDs.
|
inline |
This function pauses audio processing for a given device. Any bound audio streams will not progress, and no audio will be generated. Pausing one device does not prevent other unpaused devices from running.
Unlike in SDL2, audio devices start in an unpaused state, since an app has to bind a stream before any audio will flow. Pausing a paused device is a legal no-op.
Pausing a device can be useful to halt all audio without unbinding all the audio streams. This might be useful while a game is paused, or a level is loading, etc.
Physical devices can not be paused or unpaused, only logical devices created through AudioDeviceBase.AudioDeviceBase() can be.
Error | on failure. |
|
inline |
Unlike in SDL2, audio devices start in an unpaused state, since an app has to bind a stream before any audio will flow.
Physical devices can not be paused or unpaused, only logical devices created through AudioDeviceBase.AudioDeviceBase() can be. Physical and invalid device IDs will report themselves as unpaused here.
|
inline |
This function unpauses audio processing for a given device that has previously been paused with AudioDeviceBase.Pause(). Once unpaused, any bound audio streams will begin to progress again, and audio can be generated.
Unlike in SDL2, audio devices start in an unpaused state, since an app has to bind a stream before any audio will flow. Unpausing an unpaused device is a legal no-op.
Physical devices can not be paused or unpaused, only logical devices created through AudioDeviceBase.AudioDeviceBase() can be.
Error | on failure. |
|
inline |
The gain of a device is its volume; a larger gain means a louder output, with a gain of zero being silence.
Audio devices default to a gain of 1.0f (no change in output).
Physical devices may not have their gain changed, only logical devices, and this function will always return false when used on physical devices. While it might seem attractive to adjust several logical devices at once in this way, it would allow an app or library to interfere with another portion of the program's otherwise-isolated devices.
This is applied, along with any per-audiostream gain, during playback to the hardware, and can be continuously changed to create various effects. On recording devices, this will adjust the gain before passing the data into an audiostream; that recording audiostream can then adjust its gain further when outputting the data elsewhere, if it likes, but that second gain is not applied until the data leaves the audiostream again.
gain | the gain. 1.0f is no change, 0.0f is silence. |
Error | on failure. |
|
inline |
This is useful for accessing the final mix, perhaps for writing a visualizer or applying a final effect to the audio data before playback.
The buffer is the final mix of all bound audio streams on an opened device; this callback will fire regularly for any device that is both opened and unpaused. If there is no new data to mix, either because no streams are bound to the device or all the streams are empty, this callback will still fire with the entire buffer set to silence.
This callback is allowed to make changes to the data; the contents of the buffer after this call is what is ultimately passed along to the hardware.
The callback is always provided the data in float format (values from -1.0f to 1.0f), but the number of channels or sample rate may be different than the format the app requested when opening the device; SDL might have had to manage a conversion behind the scenes, or the playback might have jumped to new physical hardware when a system default changed, etc. These details may change between calls. Accordingly, the size of the buffer might change between calls as well.
This callback can run at any time, and from any thread; if you need to serialize access to your app's data, you should provide and use a mutex or other synchronization device.
All of this to say: there are specific needs this callback can fulfill, but it is not the simplest interface. Apps should generally provide audio in their preferred format through an AudioStreamBase and let SDL handle the difference.
This function is extremely time-sensitive; the callback should do the least amount of work possible and return as quickly as it can. The longer the callback runs, the higher the risk of audio dropouts or other problems.
This function will block until the audio device is in between iterations, so any existing callback that might be running will finish before this function sets the new callback and returns.
Setting a nullptr callback function disables any previously-set callback.
callback | a callback function to be called. Can be nullptr. |
userdata | app-controlled pointer passed to callback. Can be nullptr. |
Error | on failure. |
|
inline |
This is useful for accessing the final mix, perhaps for writing a visualizer or applying a final effect to the audio data before playback.
The buffer is the final mix of all bound audio streams on an opened device; this callback will fire regularly for any device that is both opened and unpaused. If there is no new data to mix, either because no streams are bound to the device or all the streams are empty, this callback will still fire with the entire buffer set to silence.
This callback is allowed to make changes to the data; the contents of the buffer after this call is what is ultimately passed along to the hardware.
The callback is always provided the data in float format (values from -1.0f to 1.0f), but the number of channels or sample rate may be different than the format the app requested when opening the device; SDL might have had to manage a conversion behind the scenes, or the playback might have jumped to new physical hardware when a system default changed, etc. These details may change between calls. Accordingly, the size of the buffer might change between calls as well.
This callback can run at any time, and from any thread; if you need to serialize access to your app's data, you should provide and use a mutex or other synchronization device.
All of this to say: there are specific needs this callback can fulfill, but it is not the simplest interface. Apps should generally provide audio in their preferred format through an AudioStreamBase and let SDL handle the difference.
This function is extremely time-sensitive; the callback should do the least amount of work possible and return as quickly as it can. The longer the callback runs, the higher the risk of audio dropouts or other problems.
This function will block until the audio device is in between iterations, so any existing callback that might be running will finish before this function sets the new callback and returns.
Setting a nullptr callback function disables any previously-set callback.
callback | a callback function to be called. Can be nullptr. |
Error | on failure. |