The question the researchers set out to answer is deceptively simple, yet it touches one of the deepest issues in visual neuroscience. In the natural world, objects rarely advertise a single identifying feature. A ripe berry is both red and round; a predator’s silhouette is both dark and shaped in a particular way. When multiple cues redundantly point to the same answer, an animal could in principle combine them all, weight them equally, or shortcut the whole computation by relying on whichever cue is easiest to extract. Humans and other mammals show complex, context-dependent weighting of such redundant cues. Whether birds — whose visual system is built along an entirely different anatomical plan — follow a similar logic, or adopt a simpler dominance rule, has remained unclear.

Earlier work had already hinted that pigeons prefer color over shape when classifying visual targets. What remained unknown was whether color merely takes the lead while shape still contributes meaningful discriminative information in the background, or whether shape is effectively sidelined once color is on the scene. Disentangling these possibilities required an experimental design in which the two types of information could be separated, combined, and compared across the same animals — and, crucially, a way to watch the brain respond as each kind of cue arrived.

The team’s approach was a transfer paradigm built around single-cue and multi-cue discriminations. Pigeons were first trained on tasks in which only one type of information was useful: some birds learned to classify images using color alone, while others had to rely on shape alone. Once these single-cue habits were firmly established, the birds were shifted to multi-cue tasks in which color and shape appeared together as redundant signals, both pointing toward the same classification response. This transition is the heart of the experiment. If a pigeon trained on color alone suddenly performs even better when shape is added, that suggests shape is being recruited as extra evidence. If performance stays flat, color was already doing the heavy lifting. Conversely, a shape-trained bird that improves when color joins the display would reveal that color carries information the bird can exploit immediately.

The behavioral results were asymmetric in a way that speaks volumes. Moving pigeons from the single-color task to the combined color-plus-shape task produced no measurable change in recognition accuracy — the birds were already performing at a level that the additional shape information could not improve. Moving them from the single-shape task to the multi-cue task, however, produced a significant boost in accuracy. The pattern indicates that color served as the primary behavioral driver of object recognition in these tasks, while shape contributed little detectable benefit once color was present. In other words, the pigeons did not behave like statisticians averaging two independent sources of evidence; they behaved like animals that had found one sufficient answer and saw little reason to look further.

What makes the study more than a behavioral curiosity is its second, neurophysiological layer. While the birds performed the tasks, the researchers recorded local field potentials — the summed electrical activity of populations of neurons — from the entopallium, the bird brain’s principal station for processing visual object information. The entopallium is the avian counterpart of structures in the mammalian visual pathway that support object recognition, making it the natural place to look for a neural signature of cue preference. If pigeons genuinely privilege color, that bias should leave traces in the strength and coordination of the entopallial response.

It did. At most of the recording sites, the local field potential responses were stronger for color cues than for shape cues, a neural asymmetry that mirrors the behavioral hierarchy. Even more intriguing was what happened to functional connectivity — the degree to which activity at different sites rises and falls together. When pigeons viewed multi-cue images, the local functional connectivity within the entopallium was less dense than when they viewed single-cue images. This counterintuitive finding suggests that when one dominant cue is sufficient, the brain may actually relax the coordinated communication that a harder, multi-feature computation would demand. Rather than integrating two streams of information and paying the neural cost of doing so, the color-dominant system appears to streamline its activity — an economy of processing that fits the idea that redundancy invites efficiency rather than elaboration.

The authors interpret this color preference as a potentially efficient strategy: when one cue is enough to classify an object accurately, relying primarily on that cue saves computational effort without sacrificing performance. There is an appealing ecological logic to this. Color in nature is often a fast, high-contrast signal — ripe versus unripe fruit, warning coloration, the flush of blood in a threat display — and a foraging or predator-avoiding bird may benefit from a visual system that locks onto such signals quickly. The new data add a mechanistic dimension to that intuition, showing that the preference is not just a habit of responding but is reflected in the raw amplitude of the brain’s visual response and in the architecture of its local communication.

The study also carries broader implications for how scientists think about avian intelligence. Birds achieve object recognition with a forebrain organization radically different from the mammalian cortex, yet they solve visual classification problems with a flexibility that sometimes rivals primates. Findings like these show that the strategies birds bring to such problems are not simply scaled-down versions of mammalian ones; they can exhibit their own characteristic biases, shaped by the statistics of natural scenes and the economics of neural processing. Understanding when and why a brain chooses to ignore redundant information — rather than exploit it — is a question that reaches beyond pigeons, informing models of attention, cue integration, and efficient coding in any visual system, including our own and those of artificial agents designed to mimic them.

Funded by the National Natural Science Funds of China, the China Postdoctoral Science Foundation, and the Science and Technology Research Project of Henan Province, the work is published open access, allowing researchers anywhere to examine the full behavioral and electrophysiological evidence. For now, the takeaway is vivid and a little humbling: the pigeon pecking at a screen in a laboratory is not weighing every scrap of information available to it. It has made up its mind about what matters — and in the world of these birds, at least under these conditions, color rules.

Subject of Research: Color-dominant visual object recognition and cue redundancy processing in pigeons

Article Title: Color-dominant processing in pigeon object recognition: behavioral and entopallial evidence from redundant-cue transfer tasks

Article References: Niu, X., Liu, R., Zhu, M., Peng, Y., Wang, Q., & Zhang, X. (2026). Color-dominant processing in pigeon object recognition: behavioral and entopallial evidence from redundant-cue transfer tasks. Animal Cognition. https://doi.org/10.1007/s10071-026-02113-w

Image Credits: AI Generated

DOI: 10.1007/s10071-026-02113-w

Keywords: pigeons, object recognition, color perception, shape perception, redundant visual cues, entopallium, local field potentials, animal cognition, visual neuroscience, cue transfer tasks, bird vision, functional connectivity

To understand the problem, it helps to look at how displays create color in the first place. Every television, monitor, and cinema projector builds its images from three primary light sources: red, green, and blue. By mixing these primaries in different proportions, the device can reproduce a broad spectrum of colors. On older displays, the primary emissions were relatively broad in their spectral output, which meant that even though individual viewers differed in their color vision, those differences tended to average out and the colors looked broadly similar to most people. The technology was, in a sense, forgiving of human variability.

Modern wide-gamut displays are far less forgiving. To achieve more vivid and saturated colors, manufacturers have turned to narrowband primary sources, which emit light concentrated in a very tight band of wavelengths. This produces a larger color gamut, meaning the display can show a wider range of colors, but it also amplifies the consequences of individual differences in color vision. When the light sources are narrow, small shifts in a viewer’s visual sensitivity translate into noticeable shifts in perceived color. Two people watching the same calibrated screen may therefore see subtly, and sometimes not so subtly, different images.

Andy Rider of the Institute of Ophthalmology at University College London, one of the researchers behind the new model, framed the challenge in terms of the people who make the content. The people who produce films and television shows want their work to look the same whatever device it is being viewed on, a goal that is complicated both by differences in people’s color vision and by outdated calibration methods. The new model takes into account individual variability in color vision and can be used to determine how best to adjust the three primary colors of a specific display to provide the best viewing experience for as many people as possible.

At the heart of the research is a physiological model of human color vision that captures how perception actually differs between individuals. One of the variations the researchers modeled is deuteranomaly, a common form of red-green color vision deficiency that affects roughly one in twenty men. In the model, this condition is represented by shifting the wavelengths detected by the eye’s cone cells, the photoreceptors responsible for color vision. By simulating how those shifted sensitivities interact with the narrowband light emitted by modern displays, the model can predict where colors will appear mismatched for affected viewers.

The model also accounts for changes in the eye’s lens. The lens filters out a portion of short-wavelength light before it reaches the retina, and the amount of that filtering varies both with age and from person to person. By varying the short-wavelength filtering in their simulations, the researchers could represent how the same display might appear to a younger viewer compared with an older one, or to two people of the same age whose lenses differ naturally. These physiological parameters, combined with the spectral properties of the display primaries, form the basis of the model’s predictions.

A crucial part of the work involves bridging two worlds that rarely meet: the physiology of vision and the engineering of display calibration. Rider explained that the researchers combined their realistic model of human color vision with displays that are usually calibrated and analyzed in a more engineering-based framework. Calibration is normally performed with instruments such as photometers and colorimeters, which measure light output but poorly simulate its effect on human vision. The new model helps improve those simulations, offering a way to predict not just what light a display emits but what a range of actual human observers will perceive.

To test the model’s usefulness, the researchers applied it to sixteen different displays, including consumer televisions, professional monitors used in film production, and cinema projectors. The analysis revealed that some displays exhibit much larger color discrepancies between observers than others. In other words, the choice of display technology and primary spectra directly determines how much a given screen’s colors will diverge across a population of viewers. This finding gives manufacturers a concrete, quantitative tool for evaluating whether a particular design will deliver consistent color to a broad audience or only to a narrow slice of it.

The model also yielded a practical insight about which parts of the display spectrum matter most for which viewers. The spectrum of the blue primary light strongly influences how displays are seen by people of different ages, because of its interaction with the lens’s short-wavelength filtering. Meanwhile, the spectra of the red and green primaries are what drive mismatches for people with different kinds of red-green color vision deficiency. This separation suggests a clear path forward: by changing the spectra of the primary light sources, engineers could reduce the mismatches seen by different observers without sacrificing the wide color gamut that makes modern displays appealing.

The implications extend beyond screens. Rider noted that although the research focused on displays, the modeling approach can also be used to examine similar color discrepancies, such as those arising from energy-efficient LED lighting compared with traditional incandescent lightbulbs, or in other color applications where dyes and pigments change how objects reflect or absorb light. Any situation in which a light source’s spectrum interacts with the variability of human vision could benefit from the same analysis, from architectural lighting to printed materials and signaling systems.

The researchers are careful to note the limits of the current work. Their model is based on a physiological framework of color vision that draws on their own research, but the model’s predictions have not yet been verified with real human observers. Validation is the next step, and the team plans to use what they learn to guide the development of displays that produce a wide gamut of vivid colors with better color consistency across viewers. One possibility they raise is the use of more than three primary light sources in a display, which would allow finer adjustments to produce accurate colors for a wider range of people.

The underlying tension the research exposes is one that has been building in the display industry for years. Calibration standards in common use today were developed about a hundred years ago and, as Rider pointed out, do not align with normal color vision. This can make modern displays look wrong even when they have been calibrated to look the same, because the standard observer on which calibration is based does not represent the real diversity of human eyes. A century-old reference, designed for a different era of light sources, is now being asked to govern screens whose narrowband primaries were unimaginable when it was created.

What makes this research notable is that it turns a vague complaint, that colors look different on different screens or to different people, into a quantifiable, physiologically grounded problem with an engineering solution. By simulating how narrowband red, green, and blue primaries interact with shifted cone sensitivities and age-related lens filtering, the model identifies exactly where and why color reproduction goes wrong, and points to adjustments in primary spectra as the fix. If validated with human observers, the approach could influence the next generation of televisions, professional monitors, and cinema projectors, bringing the industry closer to the ideal that filmmakers have always pursued: a film that looks the same, in every color, for everyone watching.

Subject of Research: Individual differences in human color vision and their effect on modern wide-gamut display color perception

Article Title: New research reveals why modern display colors can look different to different people

Article References: New research reveals why modern display colors can look different to different people. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: color vision, display technology, wide color gamut, deuteranomaly, cone cells, display calibration, narrowband primaries, Optics Express, University College London, color perception, lens filtering, computational modeling

To understand the study, it helps to know how thinking about attentional control has evolved. For decades, the dominant view held that attention is a two-stage system: a fast, automatic, stimulus-driven process that gets grabbed by anything visually conspicuous, followed by a slower, goal-driven process that steers attention toward what the observer is looking for. On this account, a red item in a field of green ones—a so-called color singleton—should capture attention almost involuntarily. But starting in the mid-2010s, evidence accumulated for a third possibility: active suppression. In this framework, the brain does not merely fail to be captured; it can proactively dampen the neural signal of known distractors before they seize control. Electrophysiological work identified a signature called the distractor positivity, an event-related brain potential component that appears when a salient item is successfully inhibited rather than attended.

The twist that the new study builds on concerns what can be suppressed. Earlier research suggested people could learn to suppress a specific color if it repeatedly proved distracting. But more recent experiments using what researchers call the majority search paradigm showed something stranger. In this task, participants search for a target among many same-colored items while one differently colored distractor may appear; crucially, the distractor’s color varies unpredictably across trials. If observers could only suppress a learned color, this variable distractor should capture attention. Instead, search was actually faster on trials containing the singleton distractor—a facilitation effect interpreted as evidence that the distractor had been suppressed before it could interfere. Because participants did not know which color would appear, the suppression could not be based on first-order information such as a specific hue. It had to operate on second-order information—the fact that one item is unique—or on global salience signals.

That discovery raised the question the Chinese team set out to answer: does the degree of salience matter for this higher-order mechanism? Salience, in the technical sense, refers to how strongly an item differs from its surround—how much it pops out. A high-contrast red circle among gray ones pops out dramatically; a barely darker gray circle among light gray ones barely registers. If second-order suppression simply tags any unique item and quiets it, salience should be irrelevant. But if suppression strength depends on how loud the distraction signal is, then salience should modulate the effect. Two competing accounts had already been skirmishing in the literature: one line of work argued that suppression is most effective for low-salience distractors, while other findings showed that even highly salient distractors can be proactively suppressed, arguing against the low-salience interpretation.

The researchers designed four experiments to disentangle these accounts, and all materials, data, and preregistered hypotheses were made publicly available on the Open Science Framework—a methodological transparency that has become a hallmark of rigorous attention research. Experiment 1 used the majority search paradigm and manipulated salience through set size, the number of items in the search display. In a singleton detection framework, a unique item among four items is more conspicuous than a unique item among eight, because salience is fundamentally relative. The result was striking in its null effect: higher-order suppression was essentially equivalent across set sizes, with comparable search facilitation whether the singleton stood among few or many items. Relative salience computed this way, at least, did not change the suppression benefit.

Experiments 2 and 3 shifted to the capture-probe paradigm, a more sensitive tool for measuring where attention actually goes. On each trial, after the search display, a probe appears and participants must report it; accuracy at the distractor’s location indexes whether attention was captured by or suppressed at that spot. Critically, the team manipulated the singleton’s color contrast, creating high-salience and low-salience distractors whose specific colors remained unpredictable. The results demonstrated that participants proactively suppressed singleton distractors based purely on second-order salience information—but, remarkably, the suppression was stronger for the low-salience singletons. The brain, it seemed, muted the quiet interloper more thoroughly than the blaring one.

Could this surprising asymmetry be an artifact of a weak manipulation? Experiment 4 addressed that concern with a psychophysical oddball detection task, in which participants simply judged whether any unique item was present in the display. Performance on this task validated that the high- and low-salience singletons genuinely differed in detectability and pop-out, ruling out the possibility that the salience manipulation had failed. The experiment also added eye tracking, allowing the researchers to examine oculomotor suppression—whether the eyes themselves avoided the singleton region differently depending on its salience. The eye-movement data provided converging evidence, showing how overt orienting was shaped by the same salience manipulation, and the luminance-contrast manipulation again produced the key result: stronger suppression of low-salience singletons compared to high-salience ones.

The theoretical implications cut deep. If suppression were a simple reflexive quenching of whatever signal is loudest, high-salience distractors should attract more inhibitory resources, not fewer. The observed advantage for weak singletons instead suggests that second-order suppression operates within functional limits: extremely salient items generate such strong attention-orienting signals that the suppression mechanism, even when engaged, cannot fully neutralize them, whereas moderate items fall comfortably within the mechanism’s effective range. This fits with a broader picture in which active suppression is a genuine top-down process with finite capacity, distinct from both passive filtering and learned feature-specific inhibition. It also aligns with electrophysiological findings that failed suppression of salient stimuli precedes behavioral errors—when the mechanism is overwhelmed, capture and mistakes follow.

There are also practical echoes beyond the laboratory. Drivers filtering billboards from traffic signs, radiologists scanning medical images for a lone anomaly, air traffic controllers monitoring crowded displays—all perform searches in which distracting singletons must be ignored without knowing their exact appearance in advance. The finding that suppression efficiency varies with distractor salience suggests that interface designers might reduce distraction not only by eliminating conspicuous clutter but by understanding that moderately conspicuous elements may be the ones the visual system handles best, while the loudest items are paradoxically the hardest to silence completely.

As the authors conclude, salience level emerges as a genuine potential factor influencing attentional suppression rather than an incidental detail of experimental design. The study extends the second-order suppression account from a demonstration that unique-but-unknown distractors can be suppressed to a characterization of when that suppression succeeds most fully. Future work, the researchers and their colleagues suggest, will need to trace the neural time course of these salience-dependent effects and determine how they interact with expectations, reward history, and working memory capacity. For now, the message is clear: the brain’s quieting of distraction is neither automatic nor uniform, and in the contest between attention and interruption, the loudest distractors retain a stubborn edge.

Subject of Research: How visual salience influences second-order attentional suppression of salient distractors

Article Title: The role of singleton salience in second-order suppression

Article References: Sun, Y., Zhang, Q., Xu, Y., & Chen, S. (2026). The role of singleton salience in second-order suppression. Attention, Perception, & Psychophysics, 88(7), Article 196. https://doi.org/10.3758/s13414-026-03331-7

Image Credits: AI Generated

DOI: 10.3758/s13414-026-03331-7

Keywords: visual attention, attentional suppression, salience, visual search, singleton distractors, second-order suppression, eye movements, cognitive control, color perception, inhibition, psychophysics, proactive suppression