We have found that users will get the best experience when they explicitly start and stop the recording. Sometimes it is a matter of time to let yourself get used to the habit of executing the Record → Speak → Done sequence. If you really dislike this experience you can let the system sense the end of your speech. Go into the Settings and turn ON the "Detect End-of-Speech" setting. This setting should detect when you are done speaking but it will not work as well in very noisy environments. You also might experience cutoffs in cases where you hesitate too long when you dictate.
Have more questions? Submit a request