*Critical* Missing Voice input
# support
d
Ive been experiencing issues where the model generates input that are not spoken by the assistant. I see the model output and also the voice inuput in the Logs, but those input are never spoken by the assistant. Call ID: 2dd94ef2-8ae5-4ab1-ab9f-f9836f68229f Specific log that were not spoken: 16:01:46:105 [LOG] Voice input: April 4th at 10 30 AM with Adam Gustaf, 16:01:46:417 [LOG] Voice input: or another one at 11 15 AM. Would you like to book one of these slots? Can someone help with that issue, as this is a major problem on our side
v
Hi there! I understand you're experiencing a critical issue where voice inputs appear in your logs but aren't actually spoken by the assistant. This is definitely frustrating! 👋 I can't diagnose the exact cause of your specific Call ID without further investigation, but I can definitely help with troubleshooting steps for this type of issue. Based on the information I've found, here are several potential causes and solutions for missing voice input in Vapi: ### Voice ID and Configuration Issues One common cause is incorrect voice ID configuration. According to [Discord discussions](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1333785780447084554), this could happen if "the voice either doesn't exist in your account or you haven't added them to Vapi for your own Key." ### Voice Pipeline Configuration Check your voice pipeline configuration, particularly: - **startSpeakingPlan configuration**: Incorrect settings here can cause voice input issues according to [user reports](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1343636482874933340) - **Voice file path**: An incorrect or misspelled voice file path can prevent the system from finding the voice file ([Discord thread](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1335770702074347520)) ### Transcriber Issues There might be a "deafness bug" related to issues with the transcriber (e.g., Deepgram). According to [user discussions](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1343612796717371393), trying alternative transcribers like Whisper or Azure might help. ### Using Your Call ID for Diagnosis Your Call ID (2dd94ef2-8ae5-4ab1-ab9f-f9836f68229f) is valuable for troubleshooting. You can: 1. Use it to trace call logs in the Vapi dashboard (Logs > Calls) 2. Check for PCAP files and detailed logs related to the call
3. Investigate voice issues by providing the Call ID, timestamp, and voice sample recording to support Would you like me to provide more specific debugging techniques for any of these potential causes? Sources: - [Discord discussion on voice ID configuration](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1333785780447084554) - [Discord thread on startSpeakingPlan configuration](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1343636482874933340) - [Discord thread on voice file path issues](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1335770702074347520) - [Discord thread on transcriber issues](https://discord.com/channels/1211482211119796234/1211483291191083018/threads/1343612796717371393)
d
@User any team member can look at this?
k
Thank you for reaching out to us. We've received your message and will get back to you as soon as possible.
Hi, checking if this is resolved/solved for you?
d
not resolved.
k
looking into it.
Copy code
yaml
AI: I apologize for the interruption. Here are the available slots for your appointment tomorrow. April fourth 2025. With Adam Gustaf. We have an opening at 10 30 AM or another 1 at 11 15 AM.
Hey SaNDMAN, could you share a few other Call IDs? I was going through the shared Call ID call logs, but couldn't reach any conclusion as of now why those inputs are not being voiced out. A smaller Call ID will really help me understand what is going wrong and on whose side.