I am providing the technical details and logs from our Unreal Engine environment to help you investigate the instability.
Character ID:a82b6c24-f5b8-11f0-b4e2-42010a7be027Session ID:b14a70439d338a7c6df23aa0b2e4e7a3
We have identified a specific recurring error in our logs that appears to be triggering the response failure:
Error Message:Get Response Failed name 'stt_start_time' is not definedError Code:13
This internal variable error (stt_start_time) suggests a breakdown in the speech-to-text timing or response handling within the Convai API. I have attached the full Unreal Engine log files to this email for your team to trace.
Thank you for the update. We are experiencing significant delays in receiving audio when using ElevenLabs custom voices within Convai. The wait time is currently unacceptable for a real-time conversational AI experience. Is this a known bottleneck in the Convai-ElevenLabs integration, and what is the roadmap for reducing this latency?
Thank you! I have informed our team, and they are currently reviewing the issue. They will get back to you with further information as soon as possible. In the meantime, you may continue using other ElevenLabs voices.
yes its working fine now, One issue that is still persisting, i have made voice in elevenalbs with language urdu but in convai i can see that voice in Hindi