Graphics

The visual creation of the film is based on 3D models of Notre-Dame de Paris cathedral, which have evolved over time. From its construction to the modern era, five major periods of interior architectural modification (choir, rood screen, paving, stained glass, organ, statuary) have been identified, and accordingly, these five 3D models become the backdrop for the musical and cinematic presentation of the eleven selected works.

Inside view of the cathedral mesh in Blender.

Final render in Unreal.

Initially developed by Sunmetron for the acoustics team, these models were then refined, modified, by Dada ! Animation to make them suitable for cinematic presentation. All attempts were made to make the models historically accurate, with the aid of the scientific committee, while keeping within the research projects production budget. The goal, therefore, is to provide the audience with a sense of perspective on the evolving architecture and decorative conditions within the cathedral over time, while accepting some imperfections in the geometrical model.

Each work is the occasion for its own short film: the camera travels through the building, bathed in changing and specific light, celebrating the architectural and artistic wonders of Notre-Dame. The eleven shorts are combined into one long film, forming a unique virtual concert, in the form of an immersive film with 3D audio.

Inside view of the cathedral mesh in Blender.

Final render in Unreal.

The presence of the musicians is represented by animated particles. This artistic decision was made in order to avoid the numerous issues with creating realistic human avatars on the same level of realism as the architectural model. In this manner, the sense of the musicians is maintained through representative postures and animations which respond to the corresponding tracks in the audio recording the each musician.

Due to the varying microphone sensitivities, recording levels, and dynamics of each musicians’ recording, manual adjustments were needed to the analysed audio signal automations in order to arrive at animations that appear lively and responsive for each performer.