Photo was rejected or looks wrong
Symptom: The face capture step doesn’t accept your photo, or the avatar preview looks distorted or unrealistic. Fix: Retake the photo with the following in mind:- Make sure your face is centered and well-lit - avoid shadows across your face or strong backlight
- Look directly into the camera, not off to the side
- Keep your expression neutral - avoid extreme smiles or squinting
- Make sure no other faces are visible in the frame
- Avoid wearing hats, sunglasses, or anything that obscures your face
Voice recording failed quality assessment
Symptom: After clicking Finish Recording, Velo tells you the audio quality isn’t usable and asks you to retake it. The system may show specific feedback. Fix: Read the feedback shown on screen - it will indicate what went wrong (e.g. too much background noise, low volume, clipping). Common fixes:- Move to a quieter room and close windows and doors
- Check your microphone input level in your system settings - it should be high enough to register clearly without peaking into the red
- Stay close to the microphone and maintain a consistent distance throughout
- Speak at a natural, confident pace - don’t trail off at the end of sentences
Microphone not detected
Symptom: The voice recording step shows no microphone input or the recording is completely silent. Fix: Check that your browser has microphone permission for Velo. In Chrome, click the lock icon in the address bar → Site settings → Microphone → Allow. Refresh the page and try again. Also confirm your microphone is set as the default input device in your system audio settings.VeloTwin generation takes too long or seems stuck
Symptom: After submitting your photo and voice, the generation step runs for a very long time without completing. Fix: This step typically takes 1–3 minutes. If it’s been significantly longer, try refreshing the page. If your VeloTwin appears in your My VeloTwins list, it completed successfully. If not, try the creation process again.VeloTwin voice doesn’t sound like me
Symptom: The generated voice clone sounds flat, robotic, or doesn’t capture your natural tone. Fix: This usually happens when the training recording was too monotone or had quality issues. Delete the VeloTwin and create a new one, this time:- Read the training script with natural expressiveness - vary your pace and emphasis as you normally would
- Use a better microphone or record in a quieter environment
- Speak at your normal presenting pace, not slower than usual