The scintillating scotoma is a well-documented visual phenomenon in which patients experience sparkling, brilliant white light with vivid spectral colours, often described as shimmering, kaleidoscopic, or "gorgeously coloured." It occurs during cortical ischemia, when the cortex receives insufficient blood and its neural activity measurably declines. This paper presents a double dissociation between two clinically common forms of visual ischemia that yields a direct test of competing theories of consciousness. In retinal ischemia (amaurosis fugax), the eye fails while the visual cortex remains intact and fully active. The patient experiences darkness: a black curtain, dimming, absence of vision. In cortical ischemia (ischemic scintillation during carotid artery stenosis or dissection), part of the cortex fails, its EEG flattens, and tissue may die, while the eye remains functional. The patient experiences the opposite: brilliant, sparkling white light with spectral colours, perceived as intrinsically beautiful. The production model of consciousness, which holds that the cortex generates visual experience, predicts darkness in both cases: a failing factory produces less, not more. The filter or transmission theory (Bergson 1896; Huxley 1954; Kastrup 2019), which holds that the cortex normally constrains a broader field of awareness, predicts darkness when the retinal gateway closes but light when the cortical filter fails. Only the filter theory correctly predicts both outcomes. A further implication follows. If this brilliant light correlates with the absence of neural activity rather than its presence, then it has no neural correlate in the conventional sense. The framework of neural correlates of consciousness (NCC) requires co-occurrence between experience and neural activity. Here, the experience appears when the activity disappears. If this light has no neural correlate, it is not produced by the brain. It is revealed by the brain's silence. Clinical evidence is drawn from Klingebiel et al. (2008), who documented three weeks of continuous scintillating aura with confirmed cortical infarcts, abruptly terminated by surgical restoration of blood flow; Olesen et al. (1993), who showed that ischemia-induced aura may be more frequent than migraine-induced ischemia; and Biousse et al. (1998), who found positive visual phenomena in 23 of 41 patients with carotid dissection. The author's first-person account of an acephalgic migraine aura, experienced as sparkling white light with rainbow-coloured spectral aspects and perceived as beautiful, provides phenomenological illustration.