Open adamroach opened 2 years ago
Merging #192 (853b8fc) into master (9a6e1d0) will decrease coverage by
0.04%
. The diff coverage is0.00%
.
@@ Coverage Diff @@
## master #192 +/- ##
==========================================
- Coverage 87.36% 87.31% -0.05%
==========================================
Files 18 18
Lines 1773 1774 +1
==========================================
Hits 1549 1549
- Misses 195 196 +1
Partials 29 29
Flag | Coverage Δ | |
---|---|---|
go | 87.31% <0.00%> (-0.05%) |
:arrow_down: |
wasm | 86.30% <0.00%> (-0.05%) |
:arrow_down: |
Flags with carried forward coverage won't be shown. Click here to find out more.
Impacted Files | Coverage Δ | |
---|---|---|
packetizer.go | 77.35% <0.00%> (-1.49%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 9a6e1d0...853b8fc. Read the comment docs.
Description
Currently, the
Packetize
method has a sanity check that the payload being packetized is not of zero length; if is, then the function returns without doing anything. This interacts poorly with pion's binding to OpenH264, which simply returns payloads of zero length whenencode()
is called but a frame must be skipped for bandwidth limitation reasons.The most straightforward use of these two APIs with each other leads to a situation where an application reads an empty payload from OpenH264 and writes it into the Payloader directly. When this happens, the count of samples is not advanced, leading to incorrect timestamps in the generated RTP packets (which can cause A/V sync issues). This change simply ensures that the Packetizer Timestamp gets updated under such circumstances.