Asked

Talk V2 Delay Settings

I am experiencing an issue with synchronizing the mouth movements with the voice on my "EZ-InMoov Robot Head." I am using the "Talk V2" skill along with the "Microsoft Zira Desktop" voice and text output from "OpenAI ChatGPT." Could anyone provide suggestions for the appropriate delay values in the "Talk V2" skill settings to achieve better synchronization for this setup?


Related Hardware EZ-InMoov Robot Head
Related Control Talk Servo v2

ARC Pro

Upgrade to ARC Pro

Don't limit your robot's potential – subscribe to ARC Pro and transform it into a dynamic, intelligent machine.

PRO
Canada
#1  

To achieve better synchronization of the mouth movements with the voice on your EZ-InMoov Robot Head using the Talk servo v2 skill, you'll need to adjust the delay settings to match the timing of the Microsoft Zira Desktop voice. Here are some suggestions for the delay values you can try:

  1. Constant Delay: This delay controls how long the servo stays in the minimum position for consonants. You might start with a value around 90 ms and adjust as needed.

  2. Vowel Delay: This delay controls how long the servo stays in the maximum position for vowels. A starting value of 40 ms is typical, but you can adjust it to better match the speed of the voice.

  3. Period Delay: This delay is for pauses at periods or commas. You can start with 300 ms and adjust based on how long the pauses in the speech are.

  4. Start Delay: This delay helps synchronize the start of the servo movement with the audio playback. You might start with 460 ms and adjust to ensure the mouth starts moving as the voice begins.

To fine-tune these settings, you can:

  • Test with Short Phrases: Use short phrases to test the synchronization and make small adjustments to the delay values.
  • Observe and Adjust: Watch the mouth movements and listen to the voice. If the mouth moves too early or too late, adjust the Start Delay. If the mouth doesn't open or close at the right times during speech, adjust the Constant and Vowel Delays.
  • Iterate: It may take a few iterations to get the synchronization just right, so be patient and make incremental changes.

Once you find the right settings, you can save them in the Talk servo v2 configuration for consistent performance. If you need further assistance, feel free to ask!


I am a customer support GPT-powered AI and still learning. If you have doubts about my response, please use the support documentation to verify my answer.
PRO
Synthiam
#2  

Oh I know, it’s difficult to do. Because the timing of words goes off over different sentences - stops looking natural.

One of the suggestions I’ve seen is to use the Sound servo PC. And if music is being played out of the speaker too, it’ll react. So using the ControlCommand() to pause and unpause the sound servo PC.

using that rather than talk servo might be better. Hard to say because there’s no real easy way of doing it.