Judging Thirty Seconds While the Images Change

A question interrupts an image sequence: has half the time passed? Answering would be easy with a clock, but the experiment provides none. Meanwhile, the scenes, transitions, and visible field can vary. The test does not concern a video physically speeding up. It concerns the temporal decision a person makes during the same objective interval. This distinction determines what an appealing phrase such as visual changes shortening time actually means.

Research by Sandra Malpica and colleagues, published in 2022, examines relationships between visual features and duration judgments over seconds and minutes. Participants know in advance that they will judge time. The study therefore does not test looking back at the end of a day and finding it surprisingly brief. It tests judgments while an interval unfolds. Although its experiments show patterns pointing in a common direction, the conditions do not all provide equally strong evidence, and changing the task also changes how the outcome is measured. [1]

Two different errors in a halfway judgment

In the main experiment, 89 people viewed thirty-second image sequences through a virtual-reality headset. Forty-five experienced contrast and temporal-frequency conditions, while a separate group of 44 experienced field-of-view conditions. Not every participant tested every visual feature. A sample interval first established thirty seconds and indicated the halfway point at fifteen seconds. Questions in subsequent trials appeared at one of two times unknown to participants: 13.5 or 16.5 seconds. [1]

These points separate the directions of error. Saying that more than half has passed at 13.5 seconds overestimates elapsed time. Saying that less than half has passed at 16.5 seconds underestimates it. A visual condition increasing the latter response can also increase correct responses at the earlier sampling point. Making time seem shorter therefore does not amount to improving timing accuracy generally.

Participants were instructed not to count and to answer quickly. At the end of each trial they also indicated whether they thought their previous response had been correct. Defending an earlier response supplies information different from its objective correctness. Subjective confidence does not show that a shift in temporal judgment has disappeared; it describes how the person subsequently evaluates the decision.

Visual change is not a single control knob

The researchers selected natural indoor and outdoor scenes while excluding written, synthetic, or strongly emotional material. Contrast conditions used different low- and high-contrast photographs. Temporal-frequency conditions instead added three black frames per second to the same image sequence to create flickering presentation. Field-of-view conditions compared full panoramas with more restricted salient crops. These procedures changed different properties of the visual input rather than implementing one interchangeable intervention. [1]

Higher contrast, more frequent temporal changes, and wider fields of view produced a stronger tendency toward underestimation in the main experiment. The authors interpreted this as compression of perceived duration. However, attention allocation and an internal timing mechanism were not directly measured. Observing a behavioral pattern and establishing its neural cause are different achievements.

Selecting content to reduce emotional influences does not prove that all meaning has been removed. Different photographs in the contrast conditions mean that the content was not entirely identical. The paper also acknowledges that the magnitude of visual change was not fully isolated from the absolute value of the visual feature. The findings should therefore not be treated as a law attributable exclusively to change, purified of every other possible influence.

A shared direction is not shared evidential strength

A follow-up study tested contrast and temporal frequency on a conventional display with seven participants. The patterns pointed in the same direction as the headset study, but none of the tested factors produced a statistically significant effect. Same-direction proportions in a small sample do not independently establish display equivalence or a robust replication. The larger main experiment needs to be read alongside the limitations of its smaller follow-up. [1]

This distinction prevents the overall summary of a common trend from becoming a claim of decisive evidence in every subexperiment. Similar directions across conditions can be informative, but evidential strength still depends on sample size, measurement, and uncertainty. A nonsignificant follow-up likewise does not prove that the effect is absent.

More transitions without faster playback

The longer-duration experiment used a conventional display with 51 participants. Each person was randomly assigned to thirty seconds, one minute, three minutes, or five minutes, and all trials for that person had the same total duration. Sequences consisted of three-second video clips. In the high-change condition, those clips were divided into one-second cuts and rearranged. Transitions increased while the total content and duration were retained; playback was not accelerated. [1]

The question appeared only at 55 percent of the full interval. In a three-minute trial, that point was 99 seconds. Participants reported that halfway had not yet been reached on 25 percent of low-change trials and 52 percent of high-change trials. These are response proportions. They do not mean that time became 52 percent shorter or that 52 percent of participants experienced an illusion of identical size.

The reported contrast occurred at thirty seconds, one minute, and three minutes. In the five-minute condition, the underestimation proportions were the same for both levels of visual change. This establishes that no contrast was observed at five minutes in this experiment, not a universal three-minute limit for every person and task. The authors' suggestion about fatigue is a possible interpretation rather than a cause isolated through a separate intervention.

What a later keypress means

In the final experiment, eleven participants pressed a key when they believed thirty seconds had elapsed. Both conditions contained four screens in a virtual room. Low complexity meant displaying the same video synchronously on all screens; high complexity meant displaying four different video streams. Participants also responded when something red appeared in the videos. The target duration remained fixed while the diversity of simultaneous content sources changed. [1]

Durations produced in an empty scene supplied an individual baseline. Video-condition durations were expressed relative to that personal baseline. Indicating that the target had passed took longer under high complexity; the paper reports an average 25.4 percent difference in produced duration between the video conditions. This is not objective time accelerating, nor perceived duration decreasing by precisely 25.4 percent in every person. A longer production is consistent with needing more objective time to reach the same subjective target.

This final experiment had a small sample, and the authors explicitly describe its statistical power as low. The absence of a significant difference in workload-questionnaire scores does not establish that the conditions imposed identical cognitive demands. The auxiliary task in the empty scene also differed from the video task, making it problematic to attribute every empty-to-video difference exclusively to visual complexity.

The boundary between measurement and explanation

The study's contribution is to make the possible sensitivity of duration judgments to visual presentation visible through concrete tasks. Attention moving toward visual processing offers an explanation that could reduce resources devoted to timing; it is not a uniquely established mechanism directly measured here. The reliable starting point is to ask what decision was requested. Judging whether halfway has passed and producing a target duration are different measurements. The behavioral shifts become intelligible within those measurements. The finding concerns a person's clockless judgment changing under specified conditions, not a change in how the clock itself runs.

Sources

[1] Sandra Malpica and colleagues (2022), Larger visual changes compress time: The inverted effect of asemantic visual features on interval time perception, PLOS ONE 17(3): e0265591. Open-access primary research.

Source access and factual review: 14 October 2026. Main experiments, the follow-up study, and discussion were directly read; raw data, supplementary analyses, and the experimental video were not independently reexamined. Response proportions are distinguished from duration amounts, and low-powered experiments are not presented as definitive replications. Original explanation based on the Creative Commons Attribution-licensed research, attributed to Malpica and colleagues.