Résumé

Virtual Production (VP) has rapidly evolved through the convergence of real-time rendering, motion capture, camera tracking, and immersive media systems, expanding beyond film production toward broader applications in spatial media and interactive environments. However, despite the increasing adoption of VP technologies, existing workflows remain fragmented across domains such as computer vision, photogrammetry, motion capture, and audiovisual production, limiting interoperability, reproducibility, and methodological validation within academic research infrastructures.
This article proposes a conceptual and methodological framework for developing a calibrated virtual production infrastructure oriented toward spatial media research. Rather than focusing exclusively on cinematic production pipelines, the proposed approach integrates motion capture systems, spatial acquisition processes, camera calibration procedures, and real-time rendering environments within a unified experimental ecosystem.
The research adopts a practice-based and infrastructure-embedded methodology, combining preliminary calibration studies, interoperability testing, and experimental integration between optical tracking systems, calibrated camera workflows, and real-time media platforms. Initial tests highlight recurring challenges related to spatial alignment, optical distortion, synchronization, and cross-system consistency.
Rather than presenting a closed technical solution, the article positions calibrated infrastructure design as an emerging methodological problem within spatial media research, proposing a framework for future investigation into validation protocols, integrated workflows, and reproducible VP experimentation in decentralized academic environments.