14:58:25 <RRSAgent> RRSAgent has joined #evolved-webcodecs
14:58:29 <RRSAgent> logging to https://www.w3.org/2024/09/25-evolved-webcodecs-irc
14:58:29 <tpac-breakout-bot> RRSAgent, do not leave
14:58:30 <tpac-breakout-bot> RRSAgent, make logs public
14:58:31 <tpac-breakout-bot> Meeting: Evolved Video Encoding with WebCodecs
14:58:31 <tpac-breakout-bot> Chair: Erik Språng, Eugene Zemtsov
14:58:31 <tpac-breakout-bot> Agenda: https://github.com/w3c/tpac2024-breakouts/issues/71
14:58:31 <Zakim> Zakim has joined #evolved-webcodecs
14:58:32 <tpac-breakout-bot> Zakim, clear agenda
14:58:32 <Zakim> agenda cleared
14:58:32 <tpac-breakout-bot> Zakim, agenda+ Pick a scribe
14:58:33 <Zakim> agendum 1 added
14:58:33 <tpac-breakout-bot> Zakim, agenda+ Reminders: code of conduct, health policies, recorded session policy
14:58:33 <Zakim> agendum 2 added
14:58:33 <tpac-breakout-bot> Zakim, agenda+ Goal of this session
14:58:34 <Zakim> agendum 3 added
14:58:34 <tpac-breakout-bot> Zakim, agenda+ Discussion
14:58:34 <Zakim> agendum 4 added
14:58:34 <tpac-breakout-bot> Zakim, agenda+ Next steps / where discussion continues
14:58:35 <Zakim> agendum 5 added
14:58:35 <tpac-breakout-bot> tpac-breakout-bot has left #evolved-webcodecs
18:12:43 <PeterThatcher> PeterThatcher has joined #evolved-webcodecs
18:13:30 <baboba1> baboba1 has joined #evolved-webcodecs
18:13:30 <ktoumura> ktoumura has joined #evolved-webcodecs
18:13:40 <sprang> sprang has joined #evolved-webcodecs
18:15:39 <PeterThatcher> I'm taking notes
18:15:52 <PeterThatcher> [slide 1]
18:16:31 <PeterThatcher> [slide 2]
18:16:43 <PeterThatcher> [slide 3]
18:16:59 <PeterThatcher> [slide 4]
18:17:10 <Niko> Niko has joined #evolved-webcodecs
18:17:15 <PeterThatcher> [slide 5]
18:17:25 <JonathanLennox> JonathanLennox has joined #evolved-webcodecs
18:17:28 <DanilC> DanilC has joined #evolved-webcodecs
18:17:34 <jh> jh has joined #evolved-webcodecs
18:17:39 <Bernd> Bernd has joined #evolved-webcodecs
18:17:44 <riju7> riju7 has joined #evolved-webcodecs
18:17:45 <PeterThatcher> [slide 6]
18:17:46 <eugene> eugene has joined #evolved-webcodecs
18:19:01 <jib0> jib0 has joined #evolved-webcodecs
18:19:03 <PeterThatcher> [slide 7]
18:19:13 <handellm> handellm has joined #evolved-webcodecs
18:19:23 <okin> okin has joined #evolved-webcodecs
18:20:29 <padenot> padenot has joined #evolved-webcodecs
18:20:57 <PeterThatcher> [slide 8]
18:20:59 <PeterThatcher> q+
18:21:05 <MikeW> MikeW has joined #evolved-webcodecs
18:25:38 <PeterThatcher> [slide 9]
18:26:25 <chrisguttandin> chrisguttandin has joined #evolved-webcodecs
18:28:16 <PeterThatcher> [slide 11]
18:30:50 <PeterThatcher> [slide 12]
18:31:34 <baboba1> +q
18:32:08 <padenot> q+
18:32:26 <handellm> q+
18:32:48 <JonathanLennox> q+
18:34:21 <PeterThatcher> q-
18:34:37 <PeterThatcher> [slide 12]
18:34:46 <PeterThatcher> Peter: Is there an LTR example?
18:34:51 <PeterThatcher> Erik: Not yet
18:34:57 <PeterThatcher> Peter: How many buffers can I have?
18:35:12 <PeterThatcher> Erik: AV1: 8, VP9: 8, H265: 32, H264: 16, VP8: 3
18:35:17 <PeterThatcher> q+
18:37:16 <hta> hta has joined #evolved-webcodecs
18:40:32 <PeterThatcher> From many: discussion of making the API be done a different way, using EncodeVideoChunk or some other object
18:40:57 <PeterThatcher> ?:  What are the benefits vs. scalability modes
18:41:27 <PeterThatcher> Erik: If you come up with something new, you have to go through a standards body, which is painful for what would be one page of code
18:41:46 <PeterThatcher> Benard: a lot more recovery options available with this API
18:41:50 <PeterThatcher> [slide 10]
18:42:02 <PeterThatcher> ?: About the example
18:42:19 <PeterThatcher> ?: What about things that receive feedback from the other side could be provided?
18:42:23 <baboba1> +q
18:42:41 <PeterThatcher> Erik: This is just the encoder, but you could implement feedback from the other side and then decide things, such as for an LTR
18:43:02 <handellm> q- baboba
18:43:07 <handellm> q- padenot
18:43:12 <handellm> q- handellm
18:43:22 <handellm> q+ baboba
18:43:25 <PeterThatcher> Bernard: RPSI is an example of "I hold this reference", and people have implemented it for the 1:1 case
18:43:45 <PeterThatcher> Bernard: but it doesn't seem to work for a conference
18:44:11 <PeterThatcher> Bernard: Google was working on "LTF" (?)
18:46:15 <PeterThatcher> Lennox: Might be nice to say "hey this is LTR" on the decode side (?)
18:46:30 <PeterThatcher> Lennox: what is the perf on running JS on every frame?
18:46:48 <PeterThatcher> ?: Not much compared to encoding
18:46:53 <PeterThatcher> ?: You should probably use a Worker
18:47:36 <PeterThatcher> Lennox: B-frames?
18:47:41 <joachimr5> joachimr5 has joined #evolved-webcodecs
18:47:53 <PeterThatcher> Erik: Technically, you can do that if you have side channel info
18:48:10 <PeterThatcher> Erik: But the API would need an extension to be "spec compliant" B-frames
18:48:27 <PeterThatcher> ?: Can do B-frames "manually"
18:48:47 <handellm> q-
18:48:48 <JonathanLennox> q-
18:51:30 <handellm> q- PeterThatcher
18:51:31 <PeterThatcher> q-
18:51:44 <PeterThatcher> Peter: So how does this work?
18:51:45 <handellm> q- baboba
18:52:30 <PeterThatcher> ?: You use buffers[i] instead of i, but you basically have a fixed number of buffers you can save to and reference from
18:53:58 <PeterThatcher> Discussion: Easier to update JS than browsers, but still might need OS updates to get hardware encoder support (?)
18:54:36 <PeterThatcher> [slide 13]
18:56:13 <handellm> q+ Elad
18:58:07 <PeterThatcher> [slide 18]
18:59:29 <baboba1> +q
18:59:32 <handellm> ack Elad
18:59:35 <PeterThatcher> Elad: Your view seems convincing.  Did you consult privacy experts?
18:59:39 <PeterThatcher> Erik: Not yet.
18:59:54 <PeterThatcher> Bernard: Privacy experts do not know about encode/decode
19:00:31 <PeterThatcher> Bernard: Regarding choices: my problem with returning capabilities is that I formulate a query to get a "no", so I like the idea of enumerating
19:00:42 <PeterThatcher> Bernard: Maybe we block enumeration behind some permission
19:01:38 <PeterThatcher> ?: Have you looked into MediaCapabilities?
19:02:00 <PeterThatcher> Paul: Might as well extend MediaCapabilities if you're going to extend something.
19:02:07 <handellm> ack baboba
19:02:10 <PeterThatcher> Erik: Yes, we should look into that
19:03:49 <PeterThatcher> [slide 20]
19:06:14 <JonathanLennox> q+
19:06:21 <PeterThatcher> Paul: One callback or multiple callbacks?
19:06:25 <PeterThatcher> Erik: Probably serveral
19:06:26 <PeterThatcher> q+
19:06:47 <PeterThatcher> Erik: Simulcast doesn't matter; SVC has to be in the right order
19:06:57 <handellm> ack JonathanLennox
19:07:29 <PeterThatcher> Lennox: How does this work with "manual?"
19:07:41 <PeterThatcher> Erik: You would set things differently for each layer
19:07:54 <PeterThatcher> Lennox: need control per layer about quality
19:08:21 <PeterThatcher> ? VideoEncoderEncodeOptions that always has QP value, so you can set it
19:08:45 <PeterThatcher> Bernard: what about bitrate?
19:09:09 <PeterThatcher> Erik: We're going CQP route
19:09:24 <sun> sun has joined #evolved-webcodecs
19:09:31 <PeterThatcher> Erik: target bitrate is much more complex with time, timestamps, etc
19:09:41 <handellm> ack PeterThatcher
19:11:31 <PeterThatcher> Peter: How do I control inter-layer dependencies?
19:11:44 <PeterThatcher> Erik: Use the buffers from the "manual" control from before
19:11:45 <PeterThatcher> q-
19:13:46 <joachimr5> q+
19:14:05 <PeterThatcher> Peter: What is available behind the feature flag?
19:14:24 <PeterThatcher> ?: Everything from [slide 8]
19:14:44 <PeterThatcher> ?: For hardware support: have to wait
19:14:59 <PeterThatcher> ?: Anything beyond [slide 8], we don't know yet
19:15:32 <PeterThatcher> Peter: Can you upload examples/demos somewhere?
19:15:37 <PeterThatcher> Erik: Yes
19:16:22 <joachimr5> q-
19:18:49 <sun> sun has left #evolved-webcodecs
19:21:31 <handellm> generate minutes
19:21:48 <handellm> zakim generate minutes
20:06:08 <ktoumura> ktoumura has left #evolved-webcodecs