The intermediate state during adaptation produces brief perceptual distortions

What it is

Light/dark adaptation isn't a binary "either fully adapted or not adapted at all" state — there's a transition period in between, and that period itself brings a kind of brief perceptual distortion unlike either the starting or the final stable state. It's not simply "adaptation isn't finished yet, so vision is a bit unclear" — color perception and brightness perception can show systematic distortions or misjudgments during the transition that disappear on their own once adaptation fully stabilizes.

This is easy to overlook because attention usually goes to adaptation's "starting point" (the instant the sudden change happens) and its "endpoint" (the state after full adaptation), while the process in between gets treated as a plain, linear transition from worse to better, rather than a distinct state that can produce its own independent distortions.

Why it happens

The transition period produces perceptual distortions independent of the start and end points because the retina's different photoreceptor systems and different adaptation mechanisms don't switch over in perfect synchrony. During the switch from photopic to scotopic vision, for instance, the cone system and rod system adapt at different intrinsic speeds (cones fast, rods slow), and during the transition the relative contribution each system makes to the same light signal is changing dynamically — the relative weight given to color perception fluctuates along with it, which is one reason color vision behaves unstably during the transition.

Another layer is that the adaptation mechanism itself is a sum of several sub-processes with different time constants (fast neural gain adjustment, slow photopigment regeneration, each changing at its own pace). At the starting point and the endpoint, these sub-processes are either all not yet changed or all already settled at the new level, so they stay in step with each other. Only during the interval in between do the different sub-processes sit at different points on their own respective curves, and the net effect of summing them is no longer a simple linear interpolation between either final state — which is prone to producing a brief "overshoot" or misalignment in brightness, contrast, or color perception.

Studying it

The standard way to study transition-period distortion is to densify the time sampling on top of the classic light/dark adaptation paradigm: rather than measuring only the starting and ending thresholds, test stimuli (brightness matching, hue naming, detection threshold tasks) are inserted repeatedly at different points as the adaptation process unfolds, recording perceptual performance across the whole time course rather than just the two endpoints. The independent variable is time elapsed since leaving the original adaptation state (treated as a continuous variable rather than just two endpoints); the dependent variable varies by which perceptual dimension is being studied — it could be a brightness match value, a shift in reported hue, or a non-monotonic change in threshold.

Methodologically: transition-period phenomena are typically harder to measure reliably than steady-state adaptation, because they demand tighter control over exactly when the test point falls, and individual differences in adaptation speed mean the same elapsed time corresponds to a different point in the transition for different people — experimental designs usually need to calibrate each participant's adaptation speed individually before locating test time points.

Where it stops holding

• Transition-period distortion is temporary and self-limiting — it disappears on its own once adaptation fully stabilizes, and does not indicate a persistent functional abnormality in the visual system.
• The specific form and magnitude of the distortion differ by adaptation direction (getting brighter vs. getting darker) and by which perceptual dimension is involved (brightness, color, contrast) — a blanket "you might misperceive things during adaptation" doesn't cover every case; which kind of distortion might appear needs judging case by case.
• This comes more from basic vision science than from direct interface research. Extrapolating it to a specific interface scenario (the instant of a dark-mode switch, say) currently lacks dedicated measurement of interface transition-animation timing — it remains more of a reasonable mechanistic inference than a validated interface-design conclusion.

Applying it

• For scenarios requiring users to make a precise judgment in the short window right after a sudden light change (reading color-coded information, discerning detail, performing a safety-related visual check), avoid scheduling that judgment in the first short stretch right after the sudden change; give users a buffer for the transition period to pass.
• If a product needs to display important information immediately after a light/dark switch, prefer encodings that don't rely on fine color or brightness discrimination (shape, position, icons rather than subtle differences in hue or shade alone) to carry that information, reducing the risk of misjudgment from transition-period distortion.
• Verification: design a test scenario requiring a critical judgment right after a light/dark switch, have testers perform the task immediately after actually experiencing the sudden change (not from an already well-adapted state), and check whether the error rate is significantly higher than the baseline under stable lighting; if so, that scenario is indeed affected by transition-period distortion and needs its timing or encoding adjusted.

Related

• Same group: A1.07.1 Dark adaptation takes far longer than light adaptation · A1.07.2 Sudden brightness increases cause brief disability · A1.07.6 The adaptation time constant determines how long to wait for perception to stabilize after an interface switch
• Nearby: A1.05 Color vision and opponent-channel mechanisms
• Search terms: transient adaptation · adaptation transition state · rod-cone transition · perceptual distortion