Why losing some vision can make an illusion weaker
Audiovisual illusions work because the brain fuses sight and sound into one percept, and it trusts whichever sense seems more reliable at that moment. The sound-induced flash illusion is a classic example: you see one flash but hear two beeps, and the brain reports seeing two flashes. When vision is degraded, you might expect the brain to lean harder on hearing and fall for the illusion more often. What the low-vision study found was closer to the opposite: both low-vision and sighted participants showed the illusion, but the sighted controls were more susceptible [2]. The authors frame this as a Bayesian reliability story, in which the brain weights each sense by how dependable its signal is, and the illusion tracks that weighting rather than the mere presence of vision loss [2].
The amblyopia study points the same direction using a different illusion. In the McGurk effect, a person hears one syllable while watching lips form another, and the perceived sound shifts toward what the eyes suggest. Adults with amblyopia, a condition causing reduced vision in one or both eyes, showed a shift toward auditory dominance, consistent with the brain reweighting away from a less reliable visual channel [3]. So across two different illusions, the pattern is not 'vision loss means more illusion' but 'vision loss changes which sense the brain leans on.'
Does the answer hold for people with no vision at all?
The clearest evidence here concerns partial vision loss, not blindness. The low-vision study deliberately describes its participants as retaining islands of vision rather than having a complete loss, and in that group the illusion did not increase across the board [2]. The amblyopia work likewise involves people who still have usable, if degraded, vision in at least one eye [3]. Neither study tells us how the same illusions behave in someone with no light perception, so the generalization should be kept to people with residual vision.
That distinction matters because the proposed mechanism is about relative reliability. If a visual signal is weak but present, the brain can down-weight it and lean on hearing. Whether the same reweighting, or a different strategy entirely, applies when there is no visual signal to weigh is a separate question that these studies do not answer [2][3].
Illusions are not the same as everyday seeing-and-hearing tasks
A useful caution comes from distance judgment. In a study of adults with vision loss and dual sensory loss, participants judged distances using flashing lights, pulsing tones, or both together. Visual judgments stayed accurate and matched sighted controls, while auditory estimates were more error-prone for everyone and especially for people with combined vision and hearing loss [1]. Crucially, when vision and sound were combined, the vision-loss and dual-loss groups did worse than with vision alone, which the authors suggest could reflect insufficient integration or overreliance on hearing [1].
That result is about accuracy in a spatial task, not about a perceptual illusion, and it should not be read as proof that illusions are stronger or weaker in the same people. But it reinforces the broader theme: combining senses is not automatically better, and how the brain merges them depends on which channel is trustworthy for the job at hand [1]. A person can show a reduced flash illusion and still struggle when sound is added to a distance task.
What would settle the question for people with more severe vision loss
The honest state of the evidence is that susceptibility varies with the illusion and with the degree and type of vision loss, and the studies available focus on low vision and amblyopia rather than profound blindness [2][3]. Work on auditory localization in people with sensory impairments discusses how vision and hearing cue each other, but it does not test audiovisual illusions directly, so it cannot fill that gap [5]. A study of audiovisual integration in Alzheimer's disease mentions known illusions and vision impairment but does not compare people with vision loss against sighted controls, so it also leaves the generalization question open [4].
For a curious reader, the practical takeaway is that audiovisual illusions are not a fixed package that comes with having eyes. They are flexible products of how much the brain trusts each sense, which means the right question is not 'do they generalize?' but 'under what conditions, and for which illusion?' The current evidence supports a qualified yes for people with partial vision, with weaker illusions than sighted peers, and leaves complete vision loss as an open case [2][3].
Sources used in this answer
Distance judgement in individuals with vision and hearing loss
In adults with vision loss or dual sensory loss, visual distance judgments remained accurate, auditory judgments were more error-prone (especially with hearing loss), and combining vision with sound produced worse performance than vision alone.
Did you see the sound? A Bayesian assessment on cross-modal perception in low vision
Both low-vision and sighted participants experienced the sound-induced flash illusion, but sighted controls were more susceptible, consistent with illusion strength tracking visual reliability.
Audiovisual perception in adults with amblyopia: a study using the McGurk effect
Adults with amblyopia showed a shift toward auditory dominance in the McGurk effect, consistent with reweighting away from a less reliable visual signal.
Alterations of Audiovisual Integration in Alzheimer's Disease: Y. Liu et al.: Alterations of Audiovisual Integration in AD
The paper discusses audiovisual integration changes in Alzheimer's disease and mentions vision impairment, but does not compare audiovisual illusion susceptibility between people with vision loss and sighted controls.
Effect of dual sensory loss on auditory localization: implications for intervention
The paper examines auditory localization and the interplay of vision and hearing in people with sensory impairments, but does not test audiovisual illusions.