https://livekit.io logo
Join Slack
Powered by
<@U07FRRZH2GN> How to handle background noise
# ask-ai
f
@icy-policeman-64898 How to handle background noise
✅ 1
t
Thanks for reaching out, Abhishek. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
Thanks — I can help with background-noise handling. Are you asking about reducing noise for: • Live audio/video in a browser or mobile app (LiveKit SDK clients)? • Telephony/SIP calls or agent sessions? • Pre-recorded audio files or server-side processing? A couple quick suggestions depending on which you mean: • For telephony / agent sessions: enable Krisp noise cancellation (can be turned on in SIP trunk config or room/agent session options). It improves speech-to-text, recordings, and turn detection but uses extra CPU. • For SDK clients (browser/iOS/Android): use the platform’s audio processing where possible. On iOS, you can disable 
bypassVoiceProcessing
if you want the OS to apply echo/noise suppression; conversely enabling 
bypassVoiceProcessing
is needed when you want custom processing. • For low-powered devices: avoid CPU-heavy noise suppression like Krisp to prevent performance issues; prefer lighter client-side processing. • For server-side/agent pipelines: use noise-cancellation plugins in agent sessions (example pattern: select different noise_cancellation implementations based on participant kind). If you tell me which platform and whether you want a client-side, server-side, or telephony solution, I’ll give step-by-step instructions and example code/configuration for that case. Sources: Agents telephony integration | LiveKit Documentation | LiveKit Platform | Build, run, and observe voice AI agents | LiveKit | Python SDKs | livekit/livekit_composite | DeepWiki | Turns overview | LiveKit Documentation | LiveKit Platform | Build, run, and observe voice AI agents | LiveKit, with interactive elements
l
Hey @faint-gold-41196, the assistant seems to not be updated yet but we recently launched a native ai-coustics integration in LiveKit - it handles noise cancellation and primary speaker isolation - you can read more here in our docs or in LK docs here. Let me know if you try it or if you have any questions!
t
Thanks — I see your note about the new native ai‑coustics integration (background noise reduction + primary speaker isolation) and that you linked both the ai‑coustics quickstart and the LiveKit docs. I can help you get this working or answer any questions about setup, configuration, or tradeoffs. To make my guidance most useful, which of these are you trying to do? • Enable ai‑coustics speaker isolation/noise cancellation for LiveKit Cloud rooms (real‑time sessions). • Use ai‑coustics for telephony/SIP calls or agent sessions. • Compare ai‑coustics vs Krisp or other noise-cancellation options already in your pipeline. • Tune CPU/performance tradeoffs for low‑powered clients or server-side processing. • Debug an integration that’s already enabled but not producing the expected results. If you tell me the platform (browser, iOS, Android, or server/SIP) and whether you want client-side or server-side processing, I’ll provide step‑by‑step instructions and example config/code for that case (including where to enable the feature, suggested room/session options, and any performance considerations). Sources: Pricing | LiveKit | Pricing | LiveKit | Turns overview | LiveKit Documentation | Agent session | LiveKit Documentation | Agents telephony integration | LiveKit Documentation, with interactive elements
2 Views
Open in Slack
PreviousNext
Powered by