Why a loud noise warps the timing of what follows

A viewer watches a film on a television whose audio is routed through a soundbar with a small processing delay. At first the mismatch is maddening: the lips move and the voice arrives a beat late. After twenty minutes the mismatch has vanished, and the dialogue looks perfectly in sync. Then the soundbar is switched off and the audio returns to the screen speakers, now perfectly aligned. The dialogue suddenly seems early, as if the actors are speaking before their mouths move. Nothing about the film changed. The viewer’s sense of when a sound should occur has been quietly retuned to the delay, and it now misreads the truth.

What everyone sees

What everyone sees is a person adapting to a bad setup. The first complaint about lip-sync looks like a sensitive ear, and the later silence looks like acceptance. When the delay is removed and the dialogue now seems early, the viewer blames the television, the film, or their own fatigue. The experience is reported as a fact about the equipment rather than a fact about the observer. Nobody suspects that the perceived moment of a sound is not fixed, and that a few minutes of exposure to a consistent delay can move it. The adaptation feels like the world correcting itself, because the sense of simultaneity is the only clock available.

What is actually happening

The mechanism is temporal recalibration, and it was demonstrated by Hiroshi Fujisaki, Shinsuke Shimojo, Makio Kashino, and Shin’ya Nishida in a 2004 paper in Nature Neuroscience. Participants watched flashes paired with tones that were delayed by a fixed lag. After a few minutes of exposure, the same lag was judged as simultaneous: the point of perceived synchrony had shifted toward the delay. The effect is not a decision or a guess. It is a change in the underlying estimate of when the sound occurred, and it persists after the exposure ends, which is why removing the delay produces a new, opposite illusion. Jean Vroomen and Mirjam Keetels reviewed the literature in 2010 and showed that recalibration is fast, automatic, and specific: it builds up over seconds, transfers only partially between different sound frequencies or locations, and is stronger when the delay is consistent. The brain is not measuring the world; it is running a running estimate of the world, and the estimate is updated by whatever has just happened.

Why it stays hidden

It stays hidden because the recalibrated timing feels like direct perception. Nobody experiences a shifted estimate; they experience a film that is in sync, and then a film that is out of sync. The mechanism also hides because the correction is invisible in the moment it occurs. There is no sensation of the clock moving, only a slow disappearance of the annoyance, which is easily attributed to getting used to it. And because the adaptation is specific to the sound and the setting, it does not generalise in a way that would expose it: the same person can be perfectly calibrated at home and badly calibrated in a cinema, without ever noticing that the difference is in them rather than in the room.

How temporal recalibration shapes media and conversation

The effect explains why a video call feels wrong for the first minute and normal afterwards, and why a laggy connection can seem to improve without any change in the network. The brain has moved its reference point to match the delay.

It also shapes conversation. In a room with a strong echo, or on a phone line with latency, people recalibrate to the delay and then feel a jolt when the delay disappears. The discomfort of a sudden change is the signature of an estimate that has been quietly rewritten.

The evidence in numbers and laboratory tests

Fujisaki and colleagues exposed participants to a fixed audiovisual lag and measured the point at which flashes and tones were judged simultaneous. After adaptation, that point shifted toward the lag, and the shift survived the removal of the lag.

Vroomen and Keetels showed that recalibration is not a single global clock. It is at least partly specific to the stimulus: adapting to a delayed tone at one frequency does not fully transfer to another, and the effect is stronger when the delay is stable rather than jittering. The estimate is local, fast, and automatic.

When recalibration helps and when it misleads

Recalibration is useful because real environments have consistent delays. Sound travels slowly, rooms echo, and devices add latency, so a brain that kept a fixed clock would be permanently out of step with its own surroundings.

The cost is that the adapted clock can be wrong when conditions change abruptly. A musician switching between a live room and a delayed monitor, or a surgeon moving between a direct view and a video feed, can be briefly misled by a timing estimate that is still tuned to the previous setting.

Questions readers ask

What is temporal recalibration?

It is the tendency for the perceived timing of a sound to shift after exposure to a consistent delay. The brain retunes its estimate of when a sound occurred, so a lag that was once obvious comes to feel simultaneous.

Why does lip-sync look wrong at first and then fine?

Because the brain recalibrates to the delay. After a few minutes of a consistent lag, the point of perceived synchrony moves toward the delay, and the mismatch disappears from experience even though the delay is unchanged.

Why does the sound seem early when the delay is removed?

Because the estimate of when the sound occurs has been shifted toward the old delay. When the delay disappears, the recalibrated clock makes correctly timed audio feel early, which is the mirror image of the original illusion.

How long does temporal recalibration take?

It builds up over seconds to a few minutes of exposure and is strongest when the delay is consistent. The effect can persist after the delay is removed, which is why switching setups produces a sudden jolt.