https://livekit.io logo
Join Slack
Powered by
I am doing load testing in livekit server when i u...
# ask-ai
s
I am doing load testing in livekit server when i use vp9 codec with simulcast with 1020p no report produced, why? do i need to add the some thing in the livekit server configuration?
t
Thanks for reaching out, Nouman. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
Hey there! When you're using the VP9 codec in LiveKit, it works a bit differently compared to the usual simulcast codecs like H.264 or VP8. Instead of traditional simulcast layers, VP9 uses Scalable Video Coding (SVC), and the scalability is handled through the 
scalabilityMode
property rather than 
videoSimulcastLayers
. This might be why you’re not seeing any reports during your load testing. Here are a few things to keep in mind: • Make sure your LiveKit server version supports VP9 with SVC — this feature came in starting with version 1.4.2. • Set up VP9 using the 
scalabilityMode
parameter (for example, 'L3T3' means three spatial and temporal layers). • Because VP9 relies on SVC, the way reports show up during load testing might be a bit different, so not seeing traditional simulcast reports can actually be expected. • With VP9, the browser decides the best layer setup, so you won’t have as much direct control over specific resolutions like 1080p. If you’d like a hand configuring VP9 for your load tests or want to double-check your server version, just give me a shout—I’m happy to help! Sources: Codecs and more | LiveKit Docs | README.md | CHANGELOG.md | Configuring the Client SDK for Optimal Video Quality
s
what configuration is reqiured for the server to add
t
Hey there! If you’re looking to set up your server for VP9 with SVC, here’s a handy example configuration you can use:
Copy code
{
   dynacast: true,
   publishDefaults: {
     scalabilityMode: 'L3T3',
     screenshareEncoding: VideoPresets.h1080.encoding,
     videoCodecs: 'vp9',
     videoEncoding: VideoPresets.h1080.encoding
   },
   videoCaptureDefaults: {
     resolution: VideoPresets.h1080.resolution
   }
}
Just a few key things to keep in mind: • Make sure 
dynacast
is turned on to help optimize bandwidth. • Set 
scalabilityMode
to 'L3T3' to get three spatial and temporal layers. • Double-check that 
videoCodecs
is set to 'vp9'. • Adjust the resolution and encoding settings to fit what you need, like 1080p. If you want a hand applying this configuration or have any other questions, just let me know—I’m here to help!
s
i have the helm where i need to add it
?
t
I don't have the answer you're looking for. You could also try asking your question: • in one of the other Slack channels or • to https://deepwiki.com/livekit/livekit_composite which is trained on all LiveKit source code If you find the answer, please post it here to help others!
Open in Slack
PreviousNext
Powered by