What you see can change what you hear, a phenomenon called the McGurk effect.

In the McGurk effect, first described in 1976, watching a video of someone mouthing one syllable while a different syllable plays in the audio can cause listeners to perceive a third, blended sound entirely. A classic example pairs the audio “ba” with a video of a mouth forming “ga,” and most people report hearing “da,” even though neither the audio nor the video actually contains that sound.

The effect reveals that speech perception isn’t a purely auditory process; the brain automatically combines visual information from lip movements with sound to interpret speech, which is part of why it’s often easier to understand someone in a noisy room when you can see their face. The illusion is strong and remains fairly stable even when a listener knows about the effect and is trying to listen only to the audio.