Building Dichoptic VR Vision Training: What Plant Fluorescence Probes Taught Me About Signal Design This week I read about fluorescent probes that give botanists a sharper view of xylem. The key advance: faster staining, higher specificity, better signal-to-noise. It struck me how directly this maps to what we're doing with AmblyoPunch — a Meta Quest game for amblyopia (lazy eye) that uses dichoptic Gabor stimuli in a punch-dodge mechanic. In both cases, the problem is signal quality. Conventional dyes bleed; conventional patching gives the weak eye only signal, no binocular correlation. The new probes are cleaner. Our dichoptic design presents correlated structure (Gabor patches) to each eye with controlled disparity — so the brain must fuse to succeed. For devs building vision-science apps: Gabor patterns aren't arbitrary — they match V1 receptive fields. Dichoptic rendering in Unity/OpenXR requires per-eye camera offsets. The 4-stage progression maps to contrast-balance → fusion → depth → maintenance.