Added a flag --strip_parameter_set_nalus. When enabled, parameter
set NAL units, SPS/PPS for H264 and SPS/PPS/VPS for H265, are stripped
from frames when converting NAL byte stream (AnnexB stream) to NAL
unit stream, which generates avc1/hvc1; otherwise they are not
stripped, and avc3/hev1 is generated.
Parameter set NAL units should not be stripped if they are varying
in the frames otherwise the frames may fail to be decoded.
The flag is enabled by default as we don't usually see varying
SPS/PPS/VPS and it is more space efficient with them stripped.
Set --strip_parameter_set_nalus=false to disable the flag if there
are varying SPS/PPS/VPS in the frames. This addresses #206.
Change-Id: I34bde6f33069f9f77d51a510b39f58a0f0e141aa
The issue could lead to the MPD attribute mediaPresentationDuration being
wrongly generated for WebM streams with a duration longer than INT32_MAX.
The root-cause was that StreamInfo::set_duration() accepted an int instead of a
uint64_t. This seems like a pure typo, since StreamInfo already uses a uint64_t
internally for representing duration.
This CL also removes EncryptionConfig stream data type and merges it
into StreamInfo/SegmentInfo instead.
Change-Id: Idb70ce503e61d3c951225cc78b6b15c084e16dcd
- Disable subsample encryption for VP8 in ISO-BMFF
- Apply block alignment to all subsamples for VP9 in ISO-BMFF,
instead of just superframes.
Change-Id: I8dd31cc16e87abc4d538330eaff9acb0509497df
- luma_weight_l0_flag, luma_weight_l1_flag,
chroma_weight_l0_flag, chroma_weight_l1_flag were there but not set.
- The user should use the values in pred_weight_table_l0 and
pred_weight_table_l1 instead which are set.
Change-Id: Ic9c44fb113717346938a339faf074daa32d4c2d2
Also moved MediaHandler output validation to Initialize instead.
This CL also addresses #122 with consistent chunking.
Change-Id: I60c0da6d1b33421d7828bcb827d18899e71884ce
- WebVttMediaParser uses WebVttSampleConverter to generate non
overlapping media samples.
- The media samples contains ISO BMFF boxes.
- Add kCodecWebVtt to signal that the media is webvtt and
the samples will be in ISO BMFF boxes.
Change-Id: I639902cdba7b04af75428bc20622e26b8203cfb2
Before this, HLS output did not contain language information.
Now, media playlists are properly tagged with a language in the
master playlist.
b/36134267
Change-Id: I172e946dbedd096a44cb2f917b007cc004756228
- Also sets up the packaging and verify it works.
Some of the changes are temporary to get the integration going.
Change-Id: I0cf6c379d185e157808acabb9ef58ff93d4a39ae
The spec actually requires all NAL units to be escaped, so there
is no need to escape it again.
NAL byte stream is one demarcation method; NAL unit stream is another
demarcation method. Regardless of the demarcation method used, the
requirement for NAL unit is the same. See the last paragraph of
7.4.1 NAL unit semantics on the requirement on emulation prevention bytes.
Change-Id: Icc63fcc5cf965632e331f5af5f673164c7c1663a
CENCv3 recommends only encrypting video data in slice NALs. Slice
data partition NALs should not be encrypted.
In the code, differentiate is_vcl and is_video_slice. They are the
same for H265; for H264, vcl NALs include slice data partition NALs
but video_slice NALs do not.
Change-Id: I91f4bdd76d25f0eac50e39aed350ebce3f667121
- Logic for splitting up WebVTT cues into MP4 samples.
- For any cue intervals that overlap, a new sample must be created.
Change-Id: Icf7478f42d8c5790bf6f404d512c1251dd34c405
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
- Deprecated command line flags --profile and --single_segment.
'segment_template' in stream descriptors implies live profile
and non-single segment.
- Added flag --generate_static_mpd_for_live_profile to generate
static mpd for live profile; if not set, dynamic mpd will be
generated.
Close#142
Change-Id: I78879297ed118f0f246c4753a16ad125bd6b5e4f
- Added two new fields in MpdOptions: dash_profile and mpd_type
- Updated MpdBuilder to support LiveProfile with static mpd
- Command line arguments will be updated in the next CL
Issue #142
Change-Id: Ibd35e5c9e1b1ef98043392e3f04a0af4700135c1
- Add MSVS related patterns to gitignore.
- Removed encryption_modes.h from media_base.gyp. It doesn't exist.
Change-Id: I44bc86ff7e79daf6f7bc61c21a558937530d8e66
- nalu_data points to the beginning of the nalu, instead of
beginning of the slice header which excluded the nalu
header.
- nalu_size is the size of the whole nalu.
- header_bit_size calculation is corrected. The size used to exclude
emulation prevention bytes.
Change-Id: Ic00f42caeddf7b4701f3948e6ef8a3537fd7f770
- If OnEncryptionStart() is called OnMediaStart(), someone should be
calling NotifyEncrytionStart().
Fixes#169
Change-Id: I597f317d2ddeb3091b7e9d238eb9695b4cc9c69d
Three options are supported right now:
- reuse=1|0
Allow or disallow reusing UDP sockets. Default 0.
- interface=a.b.c.d
Address of the interface over which to receive UDP multicast
streams.
- timeout=microseconds
Timeout in microseconds. Default to unlimited.
A UDP url should be of the form udp://ip:port[?options], here
is an example:
udp://224.1.2.30:88?reuse=1&interface=10.11.12.13&timeout=12345
Also deprecate --udp_interface_address flag in favor udp options.
Closes#133
Change-Id: I962d35bfedc1779d67ba20ed910207c66b7c1a15
1. Allow encryption pattern of 0:0, which is treated as 1:0;
2. Remove the special handling of pattern 1:0 in pattern cryptor
which may not always be correct;
3. Allow senc/saio/saiz boxes to be absent.
Change-Id: I372e61182ec577107e39cb601c3aed80616b036d
The original code creates temporary file in the current directory,
which may not always be writable.
Change-Id: Icc278dd5db6be6206ae29defbb7423bae39cc978
- Also import the version of the protobuf bundled in packager first,
this avoids importing the installed version of the protobuf, which
may not be compatible with this script.
- Print help if no argument is supplied.
Fixes#162
Change-Id: I1b5e87ba7fb51e81b0d4dd3791aaeb37ee0fd6ce
Cues used to be generated in the end of the file; if http range
request is not supported, clients have to download the whole
file to get to the cues.
This CL updated TwoPassSingleSegmentSegmenter to write cues
after writing webm header.
This CL also updates libwebm dependency to latest.
Closes#159
Change-Id: Ic73548e1b872e6b13a37627707e7d0ff3e556877
The original code accumulates sample durations in seconds (double)
and then generates new segment / cluster by comparing the accumulated
value with configured value.
There may be loss of precision when accumulating values in double.
For example, with a GOP of 96 frames and frame rate of 24 fps; the
accumulated frame duration for 96 frames is 3.999999999999, which is
very close to 4.0 but not 4.0. Problem would arise if segment
duration is set to 4. The created segments would have a duration of
8 seconds instead of 4 seconds.
The new code accumulates sample durations in uint64_t relative to
input time scale; the configured segment duration will be converted to
timescale for comparison. This avoids loss of precision.
Change-Id: I3ae24be82a7ce45f923a6f90fea495b8b6b2e7ef
Some subtitle files use .webvtt instead of .vtt, probably because
ffmpeg uses -f webvtt when converting from some other format.
Change-Id: If88e8d2b90e5dea6751409568d5191e923dec4aa
- Also regenerated test content to insert an iframe every one
second: ffmpeg -i source.webm -auto-alt-ref 1 -g 30 out.webm
This makes sure the content is actually encrypted with one
second clear lead in end to end test.
Change-Id: I488143c148e2d6c45ba1586f6d0d835dc46db86e
- Add decrypt_config in media sample and stream.
- Update subsamples when converting NAL unit to bytestream.
Change-Id: I7b8975a453f81b22cf74bee3c9a58b7e458cbaae
1. Replace STLDeleter with unique_ptr in container, which is
supported in C++11
2. Replace string_as_array with &string[0] which is guaranteed
to work in C++11
Change-Id: I7f39c0e51fc8a3fcbb41313094a0ca6b33db7bf3
The current behavior of logging to a file in the same directory
as the binary will break if the binary does not have write access
to that directory.
Fixes Issue #139
Change-Id: I11adfaec676fc9bbef2a5caa3261889d1fba3b88
In H264, there may be multiple consecutive video slice NAL units
in the same frame. The original code assigns a new access unit
for every video slice NAL unit, which is incorrect.
Fixes#134.
Change-Id: I4d44271df48cb08867ddd02f7494fb3573af3356
Mimic iso-bmff behavior: either all the samples in a segment are encrypted
or all the samples are clear.
Change-Id: I03bdbbf5a4b690f4d87c4dceb0f8155e6fae941e
The right way to detect the presence of a keyframe in a BlockGroup
is to look for the absence of the ReferenceBlock element [1].
Closes b/30433206
[1] https://matroska.org/technical/specs/index.html#ReferenceBlock
Change-Id: I2b3c8704cbabcd5cbee38448e5528425865561a4
Also updated packager_test not to fill in output_format by default,
which will help catch these kinds of bugs.
Closes#130
Change-Id: I77157317de822d3f543f7e27d94bb4df7a1649d2
This flag is needed to workaround the Chromium bug
https://crbug.com/398130, which uses decoding timestamp
in buffered range.
Closes#113
Change-Id: Ib8f18be7165dd968bdc36c18ce29f694235c0c26
Packager should not care about this flag, whether it is supported
is up to the decoder.
Closes#126
Change-Id: I5e05d291eab01375f7f11c951b9dc93c1358c38a
- Add MuxerListener::OnEncryptionStart() for notifying that further
segments are encrypted.
b/29621230
Change-Id: I881b29d55baaf3d04e005a3b95d898071c3f272b
- Cannot include PMT for following encrypted segments in the clear lead.
Some players consider them as conflicting codecs.
- Add EXT-X-DISCONTINUITY in front of first EXT-X-KEY to notify that the
codec is switching from e.g. "avc1" to "zavc".
b/29621230
Change-Id: I45c74813630c229d66245e992eb3a5117326bb14
Before, we converted the WebM configuration to MP4 in the video
client, however this lead to a bug when fields were missing. So
now don't convert until the cluster parser so the extra info from
the stream can be added.
This also fixes a bug where the value was not printed in the warning
logs.
b/29580732
Change-Id: If0a1a4d135f98292cdaae15a11027f42d844e85d
Since the |extra_data| field contains codec-specific configuration
data, it makes sense to call it |codec_config|.
Change-Id: If9e35165a00fe82628cf931df397a8ef06505b0d
Although superframe index is not encrypted, it should still be inserted
into subsamples list (with cipher_bytes as 0).
Change-Id: I06b38f8333a22a31a71cd67e0412218dd3aaa5c3
- Add flags and stream descriptor fields for HLS.
- Remove fields from MuxerOptions. Instead pass them directly to
HlsNotifyMuxerListener.
- Rebase segment names to the master playlist's path where it makes
sense.
Change-Id: If4f54bc56ff46dc6140859a79ed66f7b99112ed7
- Tracks other than video are protected using whole-block full-sample
encryption as specified in ISO/IEC 23001-7:2016(E) 9.7, which is
equivalent to a pattern of 1:0. This is different to the non pattern
encryption counterparts.
- Also updated the code to allow the existence of other protection
schemes in the original content, which will simply be ignored.
- The internal cryptor used by AesPatternCryptor should not use
constant iv, add a DCHECK for that.
- Optimize AesPatternCryptor handling on the special pattern 1:0.
Change-Id: Idc704e7bc6b347741336f38c6d3620fc19392960
This is part of the effort to support Opus in iso-bmff #83.
Changes in this CL:
- Add support for multiple SampleGroupDescription and SampleToGroup
boxes in TrackFragment box;
- Add support for SampleGroupDescription and SampleToGroup boxes
in SampleTable box;
- Add support for AudioRollRecoveryEntry in Sample Group and 'roll'
sample grouping type;
- Also fix a bug that default length was not set correctly in
SampleGroupDescription when constant iv is used.
Change-Id: I7e31386ea6cd17a8ee5e1dca4c1a90937d51368f
- For encrypted segments, PMT specified by SAMPLE-AES specification is
generated.
- Remove time_scale_ field from TsWriter. It is not used.
- Move ContinuityCounter to a separate file.
- Add ProgramMapTableWriter and implementations.
- continuity_counter is incremented for PAT and PMT.
Change-Id: If0dd0990203cb6990f39a6832048bdc3ff08decd
This add several unit tests for EsParserH26x to test the various
NAL unit orders and access unit contents. Some tests are disabled
and will be fixed in another patch.
Change-Id: Id5e3291e22f1fe17ada2c03c42e2cdfe226abcb2
- BytesOfProtectedData in 'cens' SHALL also be 16-byte aligned
- Fix an incorrect box definition bug w.r.t to 'cbcs' and 'cens'
- Also add various protection scheme tests
Close#77Close#78
Change-Id: I63c8d8b01ce16ed60affa97ec95fc62bc2da06df
- These warning messages are not useful since the code needs to
loop through various container types to find a matching one.
Change-Id: I1d8eca563faf24e485ac71cce32e09515c174bf2
Replaces the InitializeFromH264 and InitializeFromH265 methods with
a single Initialize method that accepts a CodecType. Also moves the
CodecType enum from NaluReader to Nalu.
Change-Id: Ie33aa9c0d772a7f3c6a2e9c867e25be3edfe7828
- CBC cryptors should accept IV of size 8 bytes - it will be zero
extended to 16 bytes.
- Fixed iv() not updated problem in AesPatternCryptor.
- Replace kChainAcrossCalls with ConstantIvFlag enum flags.
Change-Id: I3fb4de0e8abbe891e6271e779373ba53f8df660d
- Sample duration wasn't scaled in TsSegmenter to calculate the current
segment's length. So segmenting length did not always respect
segment_duration.
Change-Id: I1ef4747dbd12bde7852ffcbceb971f2775aa18b3
Parsing vui_parameters in the SPS is required to be able to extract
the resolution of the video. This also adds a method to extract
the resolution to the H.265 parser.
Issue #46
Change-Id: Idf498d2afdb36a689490151a4ae6baef0b6b73f6
Now Finalize() will close all the files used by the segmenters. This
is important to allow WebMMuxer::FireOnMediaEndEvent to get the
correct size of the media file.
This also changes the WebM muxer to use 64-bit numbers for the init
and index ranges to correctly support >4GB files.
b/28194272
Change-Id: Ia84e4a4b0756f89644efea99a1a51968b22a1338
- Add hls_name and hls_group_id fields to MuxerOptions. This is used to
pass the NAME and GROUP-ID values for EXT-X-MEDIA tag to
MuxerListener.
- Change MuxerListener::OnEncryptionInfoReady() to take an
initialization vector.
- Change MuxerListener::OnNewSegment() to take segment name.
- Reworded and formatted MuxerListener comments to Doxygen style.
Issue #85
Change-Id: Iea06e68552a56ae180177ffd6ca315a7cf39456c
- crypt_byte_block and skip_byte_block for pattern-based encryption.
- constant_iv for constant iv (used by 'cbcs' protection scheme).
- Also renamed iv_size to per_sample_iv_size.
- Also changed "bool is_encrypted" to "uint8_t is_protected" to align
with CENC spec.
Issue #78
Change-Id: I2878a91e0ebe536a08a3e3109daf157fe4440e27
- Remove EncryptionMode enums
- Remove AesEncryptor::InitializeWithRandomIv, replaced with
a static function AesCryptor::GenerateRandomIv, which should
be called to generate the iv if the iv is empty.
This change is to prepare support for CBCS and CENS pattern-based
protection schemes.
Issue #77
Change-Id: Icba35089d6e451cbea7ebbf5dd5674079f206390
- FOURCC is not a mp4 only concept. It is also used elsewhere.
- RCHECK has been used outside formats/mp4.
Also updated the FOURCC enum names to match the actual fourcc code, e.g.
FOURCC_mdat for "mdat".
Change-Id: I5be02f12695882c0fb41d359305fb8ec01303a06
- Create AesCryptor class as the common base
- AesCryptor::Crypt function will serve as an Encrypt function for
encryptor and a Decrypt function for Decryptor.
Change-Id: Ie91fb14964b5091786705bf510656f40d73af160
- Simple implementation of MPEG2 TS muxer.
- This does not take MuxerListener so a manifest cannot be generated.
- Encryption is not supported.
Issue #84
Change-Id: I117b772dbbce5437398defbd564883ad758916d5
- Segmenter should not finish a segment unless the next sample is a key
frame.
- Renamed PesPacketGenerator::Finalize() to Flush().
- Use duration from the sample instead of the one copied to PesPacket.
- Remove duration field from PesPacket.
Issue #84
Change-Id: Icd90e65fd63fdeb955e7abac3473b0b54db6ac4a
- Multi-file segmenter implementation.
- The samples are passed to PesPacketGenerator to generate PesPackets.
PesPackets are passed to TsWriter to write them to file.
Issue #84
Change-Id: Ia4cae2abe2e7df46016dcdb791baaab2818aea03
The positions in the SeekHead element should be relative to the
Segment payload, not the start of the file. Also added entries for
Tracks and SegmentInfo.
Change-Id: Id692da25fffc27d78e9f1a06d061431aeb6f1e7c
- Optimize and clean up encryption and decryption code.
- Consolidate various CBC encryption/decryption schemes into a
common class.
- Make it a constructor argument whether cipher block chain is
continuous across Encrypt/Decrypt calls.
- Also align protected region size as required in CENC spec.
Issue #77
Change-Id: I533d92ada3cd80933b532b9c3a1cca105ba66f8e
Another feature that the mac build does not support. It incorrectly
uses a copy-constructor when creating new values in a std::map. This
is incorrect and should use an in-place constructor with C++11.
Change-Id: I1e822747e97eb4ae01596d63ea05058cc383cb49
This adds a preliminary H.265 parser. The parser only support
reading the slice header. This also attaches the H.265 parser to the
VideoSliceHeaderReader so H.265 frames can have unencrypted headers.
Issue #46Closes#40
Change-Id: I8b98e1f8bb04e359d9bf896fd5b2939c326b196b
The flag --include_common_pssh will add another PSSH box in addition
to the Widevine one which will contain the key IDs of all the keys.
Closes#88
Change-Id: Ic719b19747530f0e4856cfb36471a644d572a734
The new algorithm will parse NAL unit header and only starts a new
NAL unit if it is valid, otherwise it will be considered part of
the previous NAL unit.
Closes#96
Change-Id: I45f2a0f37d51841ee8345d6d0d38fcda57e0a009
KeySource is now an abstract class that is implemented in two
types WidevineKeySource and FixedKeySource. This also updates
fixed key encryption to use v1 PSSH boxes with the common system
ID.
This also updates the --pssh argument. Now it accepts full PSSH
boxes rather than data. Multiple PSSH boxes can be concatenated
together.
Issue #88
Change-Id: I4bd0290f6d8b965f7d118f075c96e0f267d7e831
- Part 1.
- Add packager command line argument "protection_scheme" to specify
protection scheme. Plumb through packager code to enable CBC
encryption/decryption.
- Add scheme type "cbc1" to sinf.
- Refactor AES encryptor and decryptor.
- Need more work in the subsample handling.
Issue #77
Change-Id: I3a9304d89adf5efbfb226b6e805a3077b6cb8c68
* EncryptionKey now contains them rather than a PSSH box.
* Outputs PSSH boxes for each entry.
* Outputs a ContentProtection element for each entry.
* Removed SystemName and UUID from KeySource.
* Removed --scheme_id_uri packager argument.
Issue #88
Change-Id: I2651784c3220fd64f5b1773fdcd70285690cf8c0
When normalizing language tags to their shortest form, we need to
split off subtags and treat them as separate components that do not
get changed.
Furthermore, MP4 can only store language tags without subtags. So
only store the main language in MP4 output.
Fixes b/27533973
Change-Id: I96049e274aae9841e321c53039ef6464a2e61574
Kongqun Yang
sylt
Rintaro Kuroiwa
Joey Parrish
Yohann Connell
Chun-da Chen
Haoming Chen
Haoming Chen
Kyle Alexander
sammirata
Thomas Inskip
Gabe Kopley
Jacob Trimble
Bei Li