| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Bug: 12782113
Change-Id: I75ee361243c5382d9288349c554deb7bed952782
|
|
|
|
|
|
|
| |
to prevent infinite loops when processing a truncated chunk
b/13130358
Change-Id: I24c46788ebc76c242da52a8f059b7577bbaebff4
|
|
|
|
|
|
|
|
| |
Don't require that there are no other boxes between mdat and moof boxes, or
that moof immediately follows moov.
Fix SAP parsing and seeking to exact start of segment.
Change-Id: I668644c5c7866dd2512f3b7f745fba438ae6e627
|
|
|
|
|
|
|
|
| |
Rewinding a fragmented mp4 file without a segment index (sidx)
after playing it to the end would crash.
b/13028840
Change-Id: I280e74364589f0acefd432a59f85a6594c009a21
|
|
|
|
| |
Change-Id: Ib4a5ab1a9e50fe9e7747a907f8b4a8c207099fdb
|
|
|
|
| |
Change-Id: Ie3bae3f037730e316d7fca12e7a3527973f752ef
|
|\
| |
| |
| | |
Change-Id: I7b1cc71057b2bd4f771e7bcf508a8c3abd6017ce
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Contains the necessary changes to make frameworks/av build and work
on a 64-bit machine.
Signed-off-by: Craig Barber <craig.barber@arm.com>
Signed-off-by: Kévin PETIT <kevin.petit@arm.com>
Signed-off-by: Ashok Bhat <ashok.bhat@arm.com>
Signed-off-by: Marcus Oakland <marcus.oakland@arm.com>
Change-Id: I725feaae50ed8eee25ca2c947cf15aee1f395c43
|
|\ \
| |/
| |
| |
| |
| |
| | |
data in 3gp files"
* commit '6146f60c87a8a9c66036325c66dea826d2e98f6f':
Added support for ID3v2 meta data in 3gp files
|
| |
| |
| |
| |
| |
| |
| |
| | |
Added support for parsing ID3v2 meta data from
the ID32 chunk in 3gp files. The priority will be
3gpp -> ID3v2 -> iTunes per field.
Change-Id: I0282ecab58e3e5fa6bd738078d562c8bb8ce00ed
|
|/
|
|
|
|
|
|
|
|
| |
Add support for streams having an audio specific configuration with the
channel_configuration equal to zero.
Add support for 6.1 and 7.1 channel configurations.
Bug 9428126
Change-Id: Iaac2516139093579c52095d4f74ae4428f8e368a
|
|
|
|
|
|
|
|
|
|
| |
For now, do not offload LD and ELD AAC decoding because there
is no way to know if it is supported by the audio DSP implementation.
The longer term fix will be to have mapMimeToAudioFormat() use the
audio object type in track metadata to refine the AAC format and the
audio HAL list supported AAC profiles.
Change-Id: Iaa88ecf3f4ae42ad48c1b42a9b007dd80eb88147
|
|
|
|
|
|
|
|
| |
Fix HE AAC SBR sampling rate reading for explicit
signaling.
Bug: 11697128.
Change-Id: Ifec0ab9d48d9293f6774ec1efd9da9445994cb7c
|
|\ |
|
| |\
| | |
| | |
| | |
| | | |
* commit '246053380a9f628405a29a055a3f1f4fba13ed5b':
Fix metadata access
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Metadata string pointers become invalid after setting more metadata,
so don't cache them.
b/11692062
Change-Id: Iaf1afb24cf53f7fa36f49ce759355693494076e5
|
| |\ \
| | |/
| | |
| | |
| | |
| | |
| | | |
klp-dev
* commit '92092b395d59e8943a461d344f617f1dc85375a3':
stagefright: limit default max-input-size for AVC
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
AVC has a compression ratio of at least 2. Account for this, as
some codecs refuse exorbitant max-input-size requests.
Change-Id: Ib68b485879ba33a12a486f2ce18fe100be977699
Signed-off-by: Lajos Molnar <lajos@google.com>
Bug: 11599364
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If the track duration was shorter than the segment duration, the calculated
encoder padding would be negative, resulting in a crash.
b/11823061
https://code.google.com/p/android/issues/detail?id=62610
Change-Id: I3989ad88caea38d212b61355c15aec13382c6116
|
|/ /
| |
| |
| |
| |
| |
| |
| | |
MPEG4Extractor::parseDrmSINF() miscalculated data offset in parsing
IPMP Descriptors. This commit makes it to take in 2 bytes which is
the size of IPMPS_Type field.
Change-Id: I42cbb6793af9d9b2f14dbfdd7a616096002793f9
|
|\ \
| |/
|/|
| |
| | |
* commit 'ff58698ec7e0f32bf9f3e1758173fe170e17270c':
Added support for 3GPP meta data atoms
|
| |
| |
| |
| |
| |
| |
| | |
3GPP metadata is not supported, only iTunes Metadata
Format is supported.
Change-Id: I78b41c88788d5a43800c936e0ac0f98424515c62
|
| |
| |
| |
| |
| |
| | |
b/8626561
Change-Id: Ibd5168282eb33d1abdc423e15a0d9aeb5a1ad687
|
| |
| |
| |
| |
| |
| | |
b/7465749
Change-Id: I5403b74a5ae428ad28b382863a09daafc400b137
|
| |
| |
| |
| | |
Change-Id: Ib6f6994228a279ee10b389515fba04516c7c42ba
|
| |
| |
| |
| |
| |
| |
| | |
This makes FragmentedMP4Extractor obsolete. It will be removed in a
separate change.
Change-Id: Ida74c07ccf84983e20a1320ee24ffc7a5c083859
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Repeated leaks detected in mediaserver context. Most chunks small but
some bigger. Adding up it was not uncommon to see 200Mb footprint.
libc memdebug allocation call stack showed the root cause in "covr"
parsing - that explains why content often was empty but sometimes big.
Change-Id: I0de6c1eeef174d2529973ff9c6a020ec3dd44c75
|
| |\ \ |
|
| | |/
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If enabling seq_scaling_matrix_present_flag = 1 the AVCC chunk can
be larger than the original buffer size. Changed to using ABuffer
instead.
Change-Id: Idacc14b45ea2634c5e608919f3ce567f23363135
|
| |/
| |
| |
| |
| |
| |
| | |
The MPEG4Extractor was parsing 8 bits instead of 16 bits when parsing
'trkn' and 'disk'. Also added support for 16 bytes size 'disk'.
Change-Id: I22b4de2ac800881884d5759776cb380917522a87
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Make sure to clear out the mean/name/data state when parsing
apple-style metadata from tracks every time we have a full set, not
just when we find an iTunSMPB set.
AAC tracks encoded from WAV by Nero tend to put in an additional apple
style metadata tag (cdec) before the iTunSMPB tag. The sequence in
the file goes something like
mean : "com.apple.iTunes"
name : "cdec"
data : "ndaudio 1.5.4.0 / -2pass -br 320000"
mean : "com.apple.iTunes"
name : "iTunSMPB"
data : " 00000000 00000A40 000000B8 <etc...>"
If the internal state was not cleared after the first set, then when
the second instance of "name" is encountered, an attempt is made to
parse the previous data entry as an iTunSMPB tag when it is actually a
cdec tag. Afterwards, mean, name and data are all cleared, and when
the second data is encountered there is no current mean or name
present, so the gapless metadata gets skipped. By clearing the
metadata state every time we have a full set of mean/name/data, we
make sure that we are always interpreting the data field as the proper
type.
Change-Id: I196e2e3f83e434f15d5ee55ae40a74a92d5a1845
|
| |
| |
| |
| |
| |
| |
| | |
Still experimental. Set property "media.stagefright.use-fragmp4" to true to
enable.
Change-Id: I210b9c5b5164b5c5eefc31309845ee881ac7db8e
|
| |
| |
| |
| |
| |
| | |
b/9900647
Change-Id: I9ea508a2685ff8adc780edd5ecec30dd1a9b0997
|
| |
| |
| |
| |
| |
| | |
not from track media header
Change-Id: I24063183f44027b999782cc9006e9a1b56e87355
|
| |
| |
| |
| |
| |
| | |
b/9175577
Change-Id: Ie159a9c9b42e6c8d9366d0ef6a607234af569e36
|
| |
| |
| |
| | |
Change-Id: I862d89c805d738db9bdf81a8f1c5b317ff968dff
|
| |
| |
| |
| |
| |
| | |
b/8626561
Change-Id: Ibd5168282eb33d1abdc423e15a0d9aeb5a1ad687
|
| |
| |
| |
| |
| |
| | |
b/7465749
Change-Id: I5403b74a5ae428ad28b382863a09daafc400b137
|
| |
| |
| |
| | |
Change-Id: Ib6f6994228a279ee10b389515fba04516c7c42ba
|
| |
| |
| |
| |
| |
| |
| | |
This makes FragmentedMP4Extractor obsolete. It will be removed in a
separate change.
Change-Id: Ida74c07ccf84983e20a1320ee24ffc7a5c083859
|
|\ \ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Repeated leaks detected in mediaserver context. Most chunks small but
some bigger. Adding up it was not uncommon to see 200Mb footprint.
libc memdebug allocation call stack showed the root cause in "covr"
parsing - that explains why content often was empty but sometimes big.
Change-Id: I0de6c1eeef174d2529973ff9c6a020ec3dd44c75
|
|\ \ \ |
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
If enabling seq_scaling_matrix_present_flag = 1 the AVCC chunk can
be larger than the original buffer size. Changed to using ABuffer
instead.
Change-Id: Idacc14b45ea2634c5e608919f3ce567f23363135
|
|/ /
| |
| |
| |
| |
| |
| | |
The MPEG4Extractor was parsing 8 bits instead of 16 bits when parsing
'trkn' and 'disk'. Also added support for 16 bytes size 'disk'.
Change-Id: I22b4de2ac800881884d5759776cb380917522a87
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Make sure to clear out the mean/name/data state when parsing
apple-style metadata from tracks every time we have a full set, not
just when we find an iTunSMPB set.
AAC tracks encoded from WAV by Nero tend to put in an additional apple
style metadata tag (cdec) before the iTunSMPB tag. The sequence in
the file goes something like
mean : "com.apple.iTunes"
name : "cdec"
data : "ndaudio 1.5.4.0 / -2pass -br 320000"
mean : "com.apple.iTunes"
name : "iTunSMPB"
data : " 00000000 00000A40 000000B8 <etc...>"
If the internal state was not cleared after the first set, then when
the second instance of "name" is encountered, an attempt is made to
parse the previous data entry as an iTunSMPB tag when it is actually a
cdec tag. Afterwards, mean, name and data are all cleared, and when
the second data is encountered there is no current mean or name
present, so the gapless metadata gets skipped. By clearing the
metadata state every time we have a full set of mean/name/data, we
make sure that we are always interpreting the data field as the proper
type.
Change-Id: I196e2e3f83e434f15d5ee55ae40a74a92d5a1845
|
|/
|
|
|
|
|
| |
Still experimental. Set property "media.stagefright.use-fragmp4" to true to
enable.
Change-Id: I210b9c5b5164b5c5eefc31309845ee881ac7db8e
|
|
|
|
| |
Change-Id: I6cd499d257d72f50a5b508bed97796a591a51506
|
|
|
|
|
| |
Change-Id: I72474c17757dba5867f55b0e99e76e9e4e32ce7b
related-to-bug: 6217289
|