Defines an interface for controlling playback of media elementary streams.
The value of the initial ticket held by the player before the first seek. The player will use this ticket value to make the first call to SbPlayerStatusFunc with kSbPlayerStateInitialized.
The value to pass into SbPlayerCreate's
duration_pts argument for cases where the duration is unknown, such as for live streams.
Well-defined value for an invalid player.
An indicator of whether the decoder can accept more samples.
The decoder is asking for one more sample.
Identify the type of side data accompanied with
SbPlayerSampleInfo, as side data may come from multiple sources.
The side data comes from the BlockAdditional data in the Matroska/Webm container, as specified in https://tools.ietf.org/id/draft-lhomme-cellar-matroska-03.html#rfc.section.7.3.399) andnd https://www.webmproject.org/docs/container/#BlockAdditionall) . The first 8 bytes of the data contains the value of BlockAddID in big endian format, followed by the content of BlockAdditional.
An indicator of the general playback state.
The player has just been initialized. It is expecting an SbPlayerSeek() call to enter the prerolling state.
The player is prerolling, collecting enough data to fill the pipeline before presentation starts. After the first preroll is completed, there should always be a video frame to render, even if the player goes back to Prerolling after a Seek.
The player is presenting media, and it is either paused or actively playing in real-time. Note that the implementation should use this state to signal that the preroll has been finished.
The player is presenting media, but it is paused at the end of the stream.
The player has been destroyed, and will send no more callbacks.
An opaque handle to an implementation-private structure representing a player.
typedef struct SbPlayerPrivate* SbPlayer
Callback to free the given sample buffer data. When more than one buffer are sent in SbPlayerWriteSample(), the implementation only has to call this callback with
sample_buffer points to the the first buffer.
typedef void(* SbPlayerDeallocateSampleFunc) (SbPlayer player, void *context, const void *sample_buffer)
Callback for decoder status updates, called in response to a call to SbPlayerSeek() or SbPlayerWriteSample(). This callback will never be called until at least one call to SbPlayerSeek has occurred.
ticket will be set to the ticket passed into the last received call to SbPlayerSeek() at the time this callback was dispatched. This is to distinguish status callbacks for interrupting seeks. These callbacks will happen on a different thread than the calling thread, and it is not valid to call SbPlayer functions from within this callback. After an update with kSbPlayerDecoderStateNeedsData, the user of the player will make at most one call to SbPlayerWriteSample() or SbPlayerWriteEndOfStream(). The player implementation should update the decoder status again after such call to notify its user to continue writing more frames.
typedef void(* SbPlayerDecoderStatusFunc) (SbPlayer player, void *context, SbMediaType type, SbPlayerDecoderState state, int ticket)
Callback for player errors, that may set a
error: indicates the error code.
message: provides specific informative diagnostic message about the error condition encountered. It is ok for the message to be an empty string or NULL if no information is available.
typedef void(* SbPlayerErrorFunc) (SbPlayer player, void *context, SbPlayerError error, const char *message)
Callback for player status updates. These callbacks will happen on a different thread than the calling thread, and it is not valid to call SbPlayer functions from within this callback.
typedef void(* SbPlayerStatusFunc) (SbPlayer player, void *context, SbPlayerState state, int ticket)
The playback related parameters to pass into SbPlayerCreate() and SbPlayerGetPreferredOutputMode().
Provides an appropriate DRM system if the media stream has encrypted portions. It will be
kSbDrmSystemInvalid if the stream does not have encrypted portions.
Contains a populated SbMediaAudioSampleInfo if
kSbMediaAudioCodecNone, the video doesn’t have an audio track.
Contains a populated SbMediaVideoSampleInfo if
kSbMediaVideoCodecNone, the video is audio only.
Selects how the decoded video frames will be output. For example,
kSbPlayerOutputModePunchOut indicates that the decoded video frames will be output to a background video layer by the platform, and
kSbPlayerOutputDecodeToTexture indicates that the decoded video frames should be made available for the application to pull via calls to SbPlayerGetCurrentFrame().
Information about the current media playback state.
The position of the playback head, as precisely as possible, in microseconds.
The known duration of the currently playing media stream, in microseconds.
The result of getStartDate for the currently playing media stream, in microseconds since the epoch of January 1, 1601 UTC.
The width of the currently displayed frame, in pixels, or 0 if not provided by this player.
The height of the currently displayed frame, in pixels, or 0 if not provided by this player.
Whether playback is currently paused.
The current player volume in [0, 1].
The number of video frames sent to the player since the creation of the player.
The number of video frames decoded but not displayed since the creation of the player.
The number of video frames that failed to be decoded since the creation of the player.
The rate of playback. The video is played back in a speed that is proportional to this. By default it is 1.0 which indicates that the playback is at normal speed. When it is greater than one, the video is played in a faster than normal speed. When it is less than one, the video is played in a slower than normal speed. Negative speeds are not supported.
Information about the samples to be written into SbPlayerWriteSample2.
const void * buffer
Points to the buffer containing the sample data.
Size of the data pointed to by
The timestamp of the sample in SbTime.
Points to an array of side data for the input, when available.
The number of side data pointed by
side_data. It should be set to 0 if there is no side data for the input.
Information about an audio sample. This value can only be used when
type is kSbMediaTypeAudio.
Information about a video sample. This value can only be used when
type is kSbMediaTypeVideo.
union SbPlayerSampleInfo::@0 @1
The DRM system related info for the media sample. This value is required for encrypted samples. Otherwise, it must be
Side data accompanied with
SbPlayerSampleInfo, it can be arbitrary binary data coming from multiple sources.
const uint8_t * data
data will remain valid until SbPlayerDeallocateSampleFunc() is called on the
SbPlayerSampleInfo::buffer the data is associated with.
The size of the data pointed by
data, in bytes.
Creates a player that will be displayed on
window for the specified
audio_codec, acquiring all resources needed to operate it, and returning an opaque handle to it. The expectation is that a new player will be created and destroyed for every playback.
This function returns the created player. Note the following:
The associated decoder of the returned player should be assumed to not be in
kSbPlayerDecoderStateNeedsData until SbPlayerSeek() has been called on it.
It is expected either that the thread that calls SbPlayerCreate is the same thread that calls the other
SbPlayer functions for that player, or that there is a mutex guarding calls into each
If there is a platform limitation on how many players can coexist simultaneously, then calls made to this function that attempt to exceed that limit must return
kSbPlayerInvalid. Multiple calls to SbPlayerCreate must not cause a crash.
window: The window that will display the player.
window can be
kSbWindowInvalid for platforms where video is only displayed on a particular window that the underlying implementation already has access to.
video_codec: The video codec used for the player. If
kSbMediaVideoCodecNone, the player is an audio-only player. If
video_codec is any other value, the player is an audio/video decoder. This can be set to
kSbMediaVideoCodecNone to play a video with only an audio track.
audio_codec: The audio codec used for the player. The caller must provide a populated
audio_sample_info if audio codec is
kSbMediaAudioCodecAac. Can be set to
kSbMediaAudioCodecNone to play a video without any audio track. In such case
audio_sample_info must be NULL.
drm_system: If the media stream has encrypted portions, then this parameter provides an appropriate DRM system, created with
SbDrmCreateSystem(). If the stream does not have encrypted portions, then
drm_system may be
audio_sample_info: Note that the caller must provide a populated
audio_sample_info if the audio codec is
audio_sample_info can be NULL. See media.h for the format of the
SbMediaAudioSampleInfo struct. Note that
audio_specific_config is a pointer and the content it points to is no longer valid after this function returns. The implementation has to make a copy of the content if it is needed after the function returns.
max_video_capabilities: This string communicates the max video capabilities required to the platform. The web app will not provide a video stream exceeding the maximums described by this parameter. Allows the platform to optimize playback pipeline for low quality video streams if it knows that it will never adapt to higher quality streams. The string uses the same format as the string passed in to SbMediaCanPlayMimeAndKeySystem(), for example, when it is set to “width=1920; height=1080; framerate=15;”, the video will never adapt to resolution higher than 1920x1080 or frame per second higher than 15 fps. When the maximums are unknown, this will be set to NULL.
SbPlayer SbPlayerCreate(SbWindow window, const SbPlayerCreationParam *creation_param, SbPlayerDeallocateSampleFunc sample_deallocate_func, SbPlayerDecoderStatusFunc decoder_status_func, SbPlayerStatusFunc player_status_func, SbPlayerErrorFunc player_error_func, void *context, SbDecodeTargetGraphicsContextProvider *context_provider)
player, freeing all associated resources.
Upon calling this method, there should be one call to the player status callback (i.e.
player_status_func used in the creation of the player) which indicates the player is destroyed. Note, the callback has to be in- flight when SbPlayerDestroyed is called.
No more other callbacks should be issued after this function returns.
It is not allowed to pass
player into any other
SbPlayer function once SbPlayerDestroy has been called on that player.
player: The player to be destroyed.
void SbPlayerDestroy(SbPlayer player)
Given a player created with the kSbPlayerOutputModeDecodeToTexture output mode, it will return a SbDecodeTarget representing the current frame to be rasterized. On GLES systems, this function must be called on a thread with an EGLContext current, and specifically the EGLContext that will be used to eventually render the frame. If this function is called with a
player object that was created with an output mode other than kSbPlayerOutputModeDecodeToTexture, kSbDecodeTargetInvalid is returned.
SbDecodeTarget SbPlayerGetCurrentFrame(SbPlayer player)
Gets a snapshot of the current player state and writes it to
out_player_info. This function may be called very frequently and is expected to be inexpensive.
player: The player about which information is being retrieved.
out_player_info: The information retrieved for the player.
void SbPlayerGetInfo2(SbPlayer player, SbPlayerInfo2 *out_player_info2)
Writes a single sample of the given media type to
player's input stream. Its data may be passed in via more than one buffers. The lifetime of
sample_drm_info (as well as member
subsample_mapping contained inside it) are not guaranteed past the call to SbPlayerWriteSample. That means that before returning, the implementation must synchronously copy any information it wants to retain from those structures.
player: The player for which the number is retrieved.
sample_type: The type of sample for which the number is retrieved. See the
SbMediaType enum in media.h.
int SbPlayerGetMaximumNumberOfSamplesPerWrite(SbPlayer player, SbMediaType sample_type)
Returns the preferred output mode of the implementation when a video described by
creation_param is played. It is assumed that it is okay to call SbPlayerCreate() with the same video described by
creation_param, with its
output_mode member replaced by the returned output mode. When the caller has no preference on the output mode, it will set
kSbPlayerOutputModeInvalid, and the implementation can return its preferred output mode based on the information contained in
creation_param. The caller can also set
creation_param->output_mode to its preferred output mode, and the implementation should return the same output mode if it is supported, otherwise the implementation should return an output mode that it is supported, as if
creation_param->output_mode is set to
kSbPlayerOutputModeInvalid prior to the call. Note that it is not the responsibility of this function to verify whether the video described by
creation_param can be played on the platform, and the implementation should try its best effort to return a valid output mode.
creation_param will never be NULL.
SbPlayerOutputMode SbPlayerGetPreferredOutputMode(const SbPlayerCreationParam *creation_param)
Returns whether the given player handle is valid.
static bool SbPlayerIsValid(SbPlayer player)
Sets the player bounds to the given graphics plane coordinates. The changes do not take effect until the next graphics frame buffer swap. The default bounds for a player is the full screen. This function is only relevant when the
player is created with the kSbPlayerOutputModePunchOut output mode, and if this is not the case then this function call can be ignored.
This function is called on every graphics frame that changes the video bounds. For example, if the video bounds are being animated, then this will be called at up to 60 Hz. Since the function could be called up to once per frame, implementors should take care to avoid related performance concerns with such frequent calls.
player: The player that is being resized.
z_index: The z-index of the player. When the bounds of multiple players are overlapped, the one with larger z-index will be rendered on top of the ones with smaller z-index.
x: The x-coordinate of the upper-left corner of the player.
y: The y-coordinate of the upper-left corner of the player.
width: The width of the player, in pixels.
height: The height of the player, in pixels.
void SbPlayerSetBounds(SbPlayer player, int z_index, int x, int y, int width, int height)
Set the playback rate of the
rate is default to 1.0 which indicates the playback is at its original speed. A
rate greater than one will make the playback faster than its original speed. For example, when
rate is 2, the video will be played at twice the speed as its original speed. A
rate less than 1.0 will make the playback slower than its original speed. When
rate is 0, the playback will be paused. The function returns true when the playback rate is set to
playback_rate or to a rate that is close to
playback_rate which the implementation supports. It returns false when the playback rate is unchanged, this can happen when
playback_rate is negative or if it is too high to support.
bool SbPlayerSetPlaybackRate(SbPlayer player, double playback_rate)
Sets the player's volume.
player: The player in which the volume is being adjusted.
volume: The new player volume. The value must be between
1.0, inclusive. A value of
0.0 means that the audio should be muted, and a value of
1.0 means that it should be played at full volume.
void SbPlayerSetVolume(SbPlayer player, double volume)
Writes a marker to
player‘s input stream of
stream_type indicating that there are no more samples for that media type for the remainder of this media stream. This marker is invalidated, along with the rest of the stream’s contents, after a call to SbPlayerSeek.
player: The player to which the marker is written.
stream_type: The type of stream for which the marker is written.
void SbPlayerWriteEndOfStream(SbPlayer player, SbMediaType stream_type)
Writes samples of the given media type to
player's input stream. The lifetime of
sample_infos, and the members of its elements like
drm_info (as well as member
subsample_mapping contained inside it) are not guaranteed past the call to SbPlayerWriteSample2. That means that before returning, the implementation must synchronously copy any information it wants to retain from those structures.
SbPlayerWriteSample2 allows writing of multiple samples in one call.
player: The player to which the sample is written.
sample_type: The type of sample being written. See the
SbMediaType enum in media.h.
sample_infos: A pointer to an array of SbPlayerSampleInfo with
number_of_sample_infos elements, each holds the data for an sample, i.e. a sequence of whole NAL Units for video, or a complete audio frame.
sample_infos cannot be assumed to live past the call into SbPlayerWriteSample2(), so it must be copied if its content will be used after SbPlayerWriteSample2() returns.
number_of_sample_infos: Specify the number of samples contained inside
sample_infos. It has to be at least one, and less than the return value of SbPlayerGetMaximumNumberOfSamplesPerWrite().
void SbPlayerWriteSample2(SbPlayer player, SbMediaType sample_type, const SbPlayerSampleInfo *sample_infos, int number_of_sample_infos)