The sound recognition device has evolved to near perfection with a less than 0.5% error rate. When switched to Chat Mode, it recognize the pauses by the speaker and translates them as "...." The day may come when keyboards are obsolete.
Notes
The same image is used for the Japanese and US releases.