@millions-hair-24956 What audio encodings are used with the OpenAI RealTime Multi-Modal Agent ?

The audio encoding used with the OpenAI Realtime Multi-Modal Agent is "pcm16". This is defined in the AudioFormat documentation. Would you like more detail on how this encoding is used or configured? Sources: - AudioFormat documentation 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.

@millions-hair-24956 I can see in the code there is a plan to support g711-ulaw and g711-alaw. Is there any evidence this is taking place?

It seems likely that there is no current support for G.711 (mulaw or alaw) in LiveKit's OpenAI Realtime Multi-Modal Agent. According to a Slack conversation, LiveKit does not natively support mulaw for egress, and it would require conversion to Opus. There is no mention of G.711 support in the provided data sources. Would you like more detail on the current audio codec support or potential workarounds? Sources: - Slack conversation on codec support 👍 0 button 👎 0 button 💡 Hint: Mention @runllm in the thread for followups.