| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Change-Id: I81f438ae444f04c12ae27ae4ef6d073033de172c
|
|
|
|
|
|
|
|
| |
Converter now supports automatic prepending of SPS/PPS to IDR frames (h264)
as well as using the encoder in "surface-input" mode.
The new features are all opt-in and should not affect existing clients.
Change-Id: I543cf1d31ba068c1a01ab4e6814ac8d817b63faa
|
|
|
|
|
|
| |
input format, allowing control over the type of encoding.
Change-Id: Iaaa1a825f447ea130e373bbd8e5dc96f2762db18
|
|
|
|
|
|
|
|
| |
pass buffer_handle_t from encoder output to HDCP encryptor input
Bug: 8968123
Change-Id: Iea8007ce568641e213fd2e3cf6947a6f7a95746c
|
|
|
|
| |
Change-Id: I889abbbe1237e1a8fdd7135cdc91a2e9728ff39b
|
|
|
|
|
|
|
|
| |
TCP RTP channel and provide feedback on the latency of arriving packets from
the sink back to the source. This information is then used to throttle
video bitrate.
Change-Id: Ic589a3cb65e4893a3ff67de947da6063d32a1c6e
|
|
|
|
| |
Change-Id: I5f9eb4f5acb624a9e5bc0087801fb5a4a9ade35c
|
|
|
|
|
|
| |
to encode video while running as a wfd source.
Change-Id: I44f7b2350c88fc5807047c61bfe594ef8fa79275
|
|
|
|
|
|
|
|
| |
encoder input buffers on shutdown. This worked fine before for an orderly
shutdown but didn't release all references in case of error.
Change-Id: I0ea3eb26da76fbeb33cadf58d237b0c68a86ac4a
related-to-bug: 7893090
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- manually prepend SPS/PPS if encoder doesn't support it
- latency improvements
- support for "our" method of optional RTP retransmission
- improvements to the wfd commandline tool for testing
- make it easier to turn on/off suspension of the video pipeline on idle
- fixes an issue where an error during encryption would cause a SEGV
- add HDCP descriptor if necessary
Squashed commit of the following:
commit 1115be0ebb3b885b4f1b7dba56761ca013d0ec4a
Author: Andreas Huber <andih@google.com>
Date: Fri Nov 9 11:32:23 2012 -0800
Better shutdown of wfd -l sessions.
Change-Id: Id898a14ae21efd3b065b00a729830063d39195a7
commit 0e7d106dfe4eb6e2640b0b66c65deaba265f7ff0
Author: Andreas Huber <andih@google.com>
Date: Thu Nov 8 16:38:55 2012 -0800
No more sending delay, create rtp packets upfront.
Change-Id: I809a225f664fdb485c7d9a49a27886601a6a26b2
commit d399e8571b77353d59afb57508dfd2a82c1ef93a
Author: Andreas Huber <andih@google.com>
Date: Thu Nov 8 14:19:43 2012 -0800
Restore AudioSource buffer size, factor out TimeSeries, make
suspending video optional.
Change-Id: Ifdfe4d447b901e714abf52856b4641d1d55a5d41
commit f8b649f0b8f917d59f4b8a2e8e6d7db61a684a78
Author: Andreas Huber <andih@google.com>
Date: Thu Nov 8 09:34:06 2012 -0800
Pull 480 frames at a time from AudioSource/AudioRecord
Change-Id: I1e215abd329faec3da026631122c0f4c800c0ac4
commit 1bc13452eb35eebbba00f5da93fa86535be5db59
Author: Andreas Huber <andih@google.com>
Date: Thu Nov 8 08:50:30 2012 -0800
fixed bitrate traffic simulation
Change-Id: Ic5efb7cbb0b5d3b4917bc77b8ba73d447249e695
commit 016cdff18e74bdc631a5679e97192645ed095aa2
Author: Andreas Huber <andih@google.com>
Date: Wed Nov 7 14:00:03 2012 -0800
resurrected "our" style of retransmission.
Change-Id: I34d757aba67428437cb39b8293a9651750ad20d9
commit 384cf1a3c8fb4ec410bdf8fa5722c298e6028f3e
Author: Andreas Huber <andih@google.com>
Date: Tue Nov 6 09:38:55 2012 -0800
Changes to make wfd work on manta.
Change-Id: I7a4e00cf16581fe2146edd1b359af195774090e4
commit 9628f24b22b35f28630d99dda3614babf51bc07e
Author: Andreas Huber <andih@google.com>
Date: Wed Nov 7 09:15:44 2012 -0800
Patch up rtp timestamps to more accurately measure network jitter.
Change-Id: I9502a4615575f97f98a215a13131a89a6ae93c6d
commit 7c891a1a24f08bbd50f55be13f7d05f43e807eb8
Author: Andreas Huber <andih@google.com>
Date: Tue Nov 6 09:37:24 2012 -0800
Additions to the "wfd" tool to create a local wfd source.
Change-Id: I99558653a70fdc703f9d13990b3ce1c4d3ae227a
Change-Id: Ia94c63fc390f597014531073485f0cfc53b3418a
|
|
|
|
|
| |
Bug: 7426218
Change-Id: Ie1517a8017bae1f9a9b6c224cd3170dfcc5fb941
|
|
|
|
|
|
|
|
|
|
|
|
| |
related-to-bug: 7248248
May decrease power usage at the cost of significantly increasing audio bitrate.
Use "adb shell setprop media.wfd.use-pcm-audio true"
to turn it on (must be done before connecting).
Change-Id: I7ebeadf3209e01522a2644948287b23d7c383c7e
|
|
|
|
|
|
|
|
| |
No more polling the encoder for work to do, the encoder instead notifies
if there's activity.
Change-Id: Ia707211b4f5c5a6e6b70d750233d204a2d6bb778
related-to-bug: 7248248
|
|
|
|
|
|
|
|
| |
or send any audio out over the network if we've seen at least one second
of silence.
Change-Id: Ic2bca4ce7d55bd246283dc669f5f5cb6106d88ea
related-to-bug: 7248248
|
|
|
|
|
| |
Change-Id: Id029a89939e53c2bd1d20e572d3975ec2795c239
related-to-bug: 7247918
|
|
|
|
| |
Change-Id: I596cf8dd61b63465437f78413186fad2be287244
|
|
|
|
|
|
|
| |
as the video encoder allocates for its input. Do this before connecting
the bufferqueue through surface flinger.
Change-Id: Ibc81db1359afe2d88fdb75b23fadae65c8a6f7cc
|
|
Change-Id: I08f17efa0c7d007e17408feb7d4fbef0a19f531a
|