Hello,
I am currently using the CloudXR Unity plugin to connect to Omniverse.
As part of this workflow, I would like to implement a speech recognition feature inside Omniverse, which requires sending microphone audio from the Meta Quest (Unity client) to the Omniverse server.
On the Unity side, I enabled sendAudio = true. However, when attempting to connect, the Omniverse server logs show the following errors:
thread_init: already added for thread
[Error] [omni.kit.xr.system.cloudxr.plugin] [CloudXR] [FrameServer] Client requested disabled feature: Recieve audio from the client.
[Error] [omni.kit.xr.system.cloudxr.plugin] [CloudXR] [FrameServer] Disconnecting the client with error: 0X425.
[Warning] [omni.kit.xr.system.cloudxr.plugin] [CloudXR] [FrameServer] Dropping client input packets as the connection is in error state
From this, it looks like the Unity client is correctly requesting microphone audio upstream, but the Omniverse CloudXR plugin rejects it as a “disabled feature.”
I have also seen references that enabling the -ra (receive audio) option on the server side should allow client-to-server microphone audio. However, it is not clear to me where this should be configured in the Omniverse workflow.
In my case, I am not running the standalone CloudXR server binary directly. Instead, I am starting CloudXR through Omniverse by enabling the XR extension and pressing Start VR inside Omniverse. Since the server process is managed internally by the Omniverse extension, I am unsure how or where to apply the -ra option in this setup.
Could you clarify:
-
Does the Omniverse CloudXR plugin currently support client-to-server audio upstream (microphone input)?
-
If so, is there any configuration key or argument exposed in Omniverse to enable
-ra, or is this feature not yet supported in the Omniverse CloudXR integration?
Any clarification would be greatly appreciated. Thank you!