15
0%
master: 0%

Ran 24 Feb 2020 06:36AM UTC

Files 469

Run time 35s

Badge

Embed ▾

Committed 24 Feb 2020 05:42AM UTC coverage: 0.0%. Remained the same

Job # COVERALLS_PARALLEL=true build_type=debug CMAKE_EFLAGS="-DCOVERAGE=ON"

Build Type

push

travis-ci

Committed by

web-flow

Commit Message

packetization_kernel: pack entire temproal unit for a output buffer (#1053)

* packetization_kernel: pack entire temproal unit for a output buffer

Currently, we output one frame one get ouptut, and the "show existing frame" shares the same pts with the previous frame.
This will introduce jitter to "show existing frame" and previous frame.
We should output a temporal unit a time.
A temporal unit includes one TD + zero or more undisplay frame + one display frame.
And we need to stamp a different pts for show existing frame since it is a displayable frame.

after apply this patch, ffmpeg will not report error for following test command:
ffmpeg -i 320x240.y4m -c:v libsvt_av1 -vframes 100 test.mkv  -y &&  ffmpeg -c:v libdav1d -i test.mkv -f null /dev/null

and ffmpeg will not insert td for every frame if we use following command:
ffmpeg -i 320x240.y4m -c:v libsvt_av1 -vframes 100 test.ivf  -y

We always output one temporal unit a time.
Previous complex code is not needed.
every output buffer contains one temproal unit, so the e2e unit test
can be more simpler

Run Details

0 of 112465 relevant lines covered (0.0%)

0.0 hits per line

OpenVisualCloud / SVT-AV1 / 4887 / 15
0%
master: 0%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 4887.15 (COVERALLS_PARALLEL=true build_type=debug CMAKE_EFLAGS="-DCOVERAGE=ON")

OpenVisualCloud / SVT-AV1 / 4887 / 15 0% master: 0%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 4887.15 (COVERALLS_PARALLEL=true build_type=debug CMAKE_EFLAGS="-DCOVERAGE=ON")

OpenVisualCloud / SVT-AV1 / 4887 / 15
0%
master: 0%

README BADGES
x