Ogg writer #50
Ogg writer #50
Conversation
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## master #50 +/- ##
==========================================
+ Coverage 87.21% 88.03% +0.81%
==========================================
Files 19 23 +4
Lines 829 894 +65
==========================================
+ Hits 723 787 +64
- Misses 106 107 +1
Continue to review full report in Codecov by Sentry.
|
| # values set to 0 are handled by PeerConnection.set_rtp | ||
| rtp_packet = OpusPayloader.payload(packet) | ||
| rtp_packet = %{rtp_packet | timestamp: state.last_audio_timestamp} | ||
| rtp_packet = %{rtp_packet | timestamp: trunc(state.last_audio_timestamp)} |
There was a problem hiding this comment.
Can this ever be non-integer? We calculate last_audio_timestamp as: last_audio_timestamp: state.last_audio_timestamp + timestamp_delta where timestamp_delta is timestamp_delta = trunc(duration * 48_000 / 1000) (so integer) and initial last_audio_timestamp is also integer
There was a problem hiding this comment.
Yeah, this is just temporary (duration from OggReader.next_packet is now a float), I'll make sure there's no unnecessary truncs in the examples PR.
| @@ -1,6 +1,6 @@ | |||
| defmodule ExWebRTC.Media.OggReader do | |||
There was a problem hiding this comment.
If we have Ogg.Header and Ogg.Packet, it might be resonable to make everything the same in this PR (including IVF modules)
There was a problem hiding this comment.
I was wondering about the same, but all of the Ogg submodules are @moduledoc false so from API'a standpoint, it's still consistent.
Adds the
OggWriterand its tests.Example will be updated in the following Pull Request.