Struct re_video::demux::VideoData

pub struct VideoData {
    pub config: Config,
    pub timescale: Timescale,
    pub duration: Time,
    pub gops: Vec<GroupOfPictures>,
    pub samples: Vec<Sample>,
    pub samples_statistics: SamplesStatistics,
    pub mp4_tracks: BTreeMap<TrackId, Option<TrackKind>>,
}

Decoded video data.

Fields

config: Config

timescale: Timescale

How many time units are there per second.

duration: Time

Duration of the video, in time units.

gops: Vec<GroupOfPictures>

We split video into GOPs, each beginning with a key frame, followed by any number of delta frames.

samples: Vec<Sample>

Samples contain the byte offsets into data for each frame.

This list is sorted in ascending order of decode timestamps.

Samples must be decoded in decode-timestamp order, and should be presented in composition-timestamp order.

In MP4, one sample is one frame.

samples_statistics: SamplesStatistics

Meta information about the samples.

mp4_tracks: BTreeMap<TrackId, Option<TrackKind>>

All the tracks in the mp4; not just the video track.

Can be nice to show in a UI.

Implementations

impl VideoData

pub fn load_mp4(bytes: &[u8]) -> Result<Self, VideoLoadError>

impl VideoData

pub fn load_from_bytes( data: &[u8], media_type: &str ) -> Result<Self, VideoLoadError>

Loads a video from the given data.

TODO(andreas, jan): This should not copy the data, but instead store slices into a shared buffer. at the very least the should be a way to extract only metadata.

pub fn duration(&self) -> Duration

Length of the video.

pub fn dimensions(&self) -> [u32; 2]

Natural width and height of the video

pub fn width(&self) -> u32

Natural width of the video.

pub fn height(&self) -> u32

Natural height of the video.

pub fn human_readable_codec_string(&self) -> String

The codec used to encode the video.

pub fn num_samples(&self) -> usize

The number of samples in the video.

pub fn subsampling_mode(&self) -> Option<ChromaSubsamplingModes>

Returns the subsampling mode of the video.

Returns None if not detected or unknown.

pub fn bit_depth(&self) -> Option<u8>

Per color component bit depth.

Usually 8, but 10 for HDR (for example).

pub fn is_monochrome(&self) -> Option<bool>

Returns None if the mp4 doesn’t specify whether the video is monochrome or we haven’t yet implemented the logic to determine this.

pub fn frame_timestamps_ns(&self) -> impl Iterator<Item = i64> + '_

Determines the video timestamps of all frames inside a video, returning raw time values.

Returned timestamps are in nanoseconds since start and are guaranteed to be monotonically increasing.

fn latest_sample_index_at_decode_timestamp( samples: &[Sample], decode_time: Time ) -> Option<usize>

For a given decode (!) timestamp, returns the index of the first sample whose decode timestamp is lesser than or equal to the given timestamp.

fn latest_sample_index_at_presentation_timestamp_internal( samples: &[Sample], sample_statistics: &SamplesStatistics, presentation_timestamp: Time ) -> Option<usize>

See Self::latest_sample_index_at_presentation_timestamp, split out for testing purposes.

pub fn latest_sample_index_at_presentation_timestamp( &self, presentation_timestamp: Time ) -> Option<usize>

For a given presentation timestamp, return the index of the first sample whose presentation timestamp is lesser than or equal to the given timestamp.

Remember that samples after (i.e. with higher index) may have a lower presentation time if the stream has sample reordering!

pub fn gop_index_containing_decode_timestamp( &self, decode_time: Time ) -> Option<usize>

For a given decode (!) timestamp, return the index of the group of pictures (GOP) index containing the given timestamp.

pub fn gop_index_containing_presentation_timestamp( &self, presentation_timestamp: Time ) -> Option<usize>

For a given presentation timestamp, return the index of the group of pictures (GOP) index containing the given timestamp.

Trait Implementations

impl Clone for VideoData

fn clone(&self) -> VideoData

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for VideoData

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.