<p>In this study, a novel audio-visual integrated modular display system is proposed to resolve the spatial mismatch between visual and auditory information in large-scale environments. A hybrid hardware structure combining transparent LED films and aluminum plates is designed to function as a direct sound radiator driven by an array of voice coil motors (VCMs). To control the localized vibration field within the complex modular display, a zone-based rendering strategy utilizing the Moore-Penrose pseudo-inverse method is developed, exploiting the overdetermined nature of the actuator-sensor configuration. Experimental validation using a 2 × 5 modular prototype demonstrates that vibration energy is distinctly concentrated within designated target areas. The experimental results further show that the localized vibration field is well represented in the low- and mid-frequency bands below 500 Hz. These results indicate that the proposed system effectively realizes high-fidelity localized sound-on-display performance.</p>