Compression of redundant visual information improves feature discrimination in human vision

Our eyes are constantly bombarded with more visual data than the brain can possibly process in detail. To cope with this flood of information, the visual system does not simply record everything like a camera; instead, it selects, integrates, and compresses what it sees. This compression is most obvious in our peripheral vision, the area of sight we use to see things out of the corner of our eye. In this region, the brain often groups nearby objects together, sometimes making it impossible to count them individually or distinguish their specific shapes. This phenomenon is known as crowding, where surrounding items blur together and hide a target. A related effect, called redundancy masking, occurs when identical items are presented together. In this case, the brain seems to decide that repeating the same object is unnecessary, so it compresses the group into fewer items than are actually there. For instance, if three identical bars are shown in the periphery, a person might genuinely perceive only two.

For a long time, scientists assumed that this kind of compression was a necessary trade-off: the brain saves energy and processing power by discarding details, but the cost is a loss of accuracy. The prevailing view was that when the visual system merges items, it simply averages their features, leading to a fuzzy or distorted perception. However, a new study challenges this idea. The researchers asked a simple but profound question: does compressing redundant information actually make our vision worse, or could it somehow make the remaining details clearer? By testing how people perceive the width of identical bars in their peripheral vision, the team discovered that when the brain "hides" a repeated item, the features of the items that remain are actually judged more accurately than when all items are seen. This suggests that the visual system is not just a passive filter that loses information, but an active editor that discards the unnecessary to sharpen the essential.

To investigate this, the researchers conducted two experiments with human participants. They placed small black bars on a gray screen and asked people to look at them from the corner of their eye, specifically at a point ten degrees away from the center of their vision. The bars were arranged in groups of three, four, or five. After seeing the bars for a brief moment, participants had to do two things. First, they reported how many bars they thought they saw. Second, they adjusted a single bar (or a group of bars in the second experiment) in the center of their vision to match the width they had perceived in the periphery. The researchers carefully controlled the size of the bars and the space between them to ensure that the total length of the group remained the same, regardless of how many bars were actually present. This prevented participants from guessing the number of bars based on the overall size of the group.

The results confirmed that redundancy masking was happening frequently. When three bars were shown, participants often reported seeing only two. When four or five bars were shown, they frequently reported seeing fewer than the actual number. This confirmed that the brain was indeed compressing the visual input by removing the perception of identical items. The critical finding emerged when the researchers looked at how accurately people judged the width of the bars. In the trials where participants saw all the bars (or reported the correct number), they consistently underestimated the width of the bars. This is a known limitation of peripheral vision, where objects often appear smaller than they really are. However, in the trials where redundancy masking occurred—where the participants reported seeing fewer bars than were actually there—their judgment of the width was remarkably accurate. In fact, when they perceived fewer bars, the width they reported was much closer to the true physical width of the bars than when they saw all of them.

The researchers also measured the spacing between the bars. They found that when participants perceived fewer bars, the space between the remaining bars felt larger than it actually was. This expansion of perceived space, combined with the accurate width judgment, meant that the overall group felt smaller than it physically was, even though the individual bars felt wider. To understand what was happening inside the brain, the team built a computer model to test different theories. One theory suggested that the brain simply pools the information from the missing bar into the remaining ones, like mixing paint to get a new color. If this were true, the missing bar's width would be added to the others, making them appear significantly wider than they were. The data ruled this out. The model showed that the brain did not simply average the features. Instead, it appeared to discard the information from the missing bar almost entirely, while simultaneously correcting the underestimation of the bars that remained.

This finding overturns the simple idea that compression always leads to a loss of detail. The study suggests that when the visual system encounters a pattern of identical items, it recognizes the redundancy and removes the extra copies from conscious awareness. But rather than leaving the remaining items distorted, this process seems to refine their representation. The brain effectively says, "I don't need to count these three identical bars; I just need to know what one of them looks like," and in doing so, it sharpens the perception of that single item. The researchers found that this improvement in accuracy came without any loss of precision; the participants were just as consistent in their judgments when they saw fewer bars as when they saw more.

The study also explored how this compression affects the overall shape of the group. Even though the space between the perceived bars felt larger, the total length of the group felt shorter. This indicates that the brain is not just changing individual features but is reorganizing the entire spatial layout of the scene. Despite these changes in how many bars were seen, how wide they were, and how far apart they seemed, the overall density of the group—the amount of black bar relative to the total space—remained surprisingly stable. This suggests that while the brain is willing to alter the specific details of individual items to improve their clarity, it maintains a stable sense of the overall texture or pattern of the scene.

These results provide a new perspective on how the visual system handles information. Rather than viewing peripheral vision as a blurry, low-resolution version of central vision that simply averages out details, this work suggests it is a sophisticated system that actively manages information. By identifying and removing redundant copies of the same object, the brain frees up resources to represent the remaining features with greater fidelity. The study does not claim that this process works for all types of objects or in all situations, but it demonstrates that for identical items, the loss of a conscious item is not a failure of vision. Instead, it is a functional strategy that allows the brain to see the world more clearly by seeing less of it. The visual system, it turns out, knows that sometimes the best way to see the whole picture is to let go of the parts that are just repeating themselves.