The time-intensity uncertainty principle in vision

The relationship between time perception and brightness perception remains poorly understood. Here we present a computational account linking the two domains, grounded in established principles of neural information processing in visual cortex. A nonlinear transducer maps luminance to population spike rate, while correlated gain fluctuations impose an upper bound on achievable signal-to-noise ratios. Perceptual magnitudes in both domains are decoded from the same spike-count statistics, yielding a reciprocal trade-off in perceptual resolution: brighter stimuli improve temporal precision but impair brightness sensitivity, whereas longer stimuli enhance brightness sensitivity but degrade temporal resolution. We tested this conjectured trade-off in two psychophysical experiments manipulating stimulus duration and luminance. Model predictions closely matched the behavioral data, revealing a fundamental coding limit in vision: the time-intensity uncertainty principle. This limit provides a unified explanation for near-miss relations to Weber's law for time perception and intensity perception, Bloch-like temporal summation effects governing brightness discrimination sensitivity, and luminance-dependent shifts in duration discrimination sensitivity.