The project Vikrant Rona came to Viga Entertainment technology with the requirement of creating a photorealistic bike sequence in a forest. To accomplish realism, the team at VigaET, choose to use the Unreal engine's potential to create highly photorealistic forests. The forests were using a pack of megascans as well as custom-created tree vegetation to match the physical set of the film that the filmmakers initially created.
A film which required real-time Digital humans to be working for a action sequence, final pixels rendered from Unreal Engine.
The requirement of digital humans in Vikrant, Rana, stemmed from the reason that the film required a stunt sequence, The digital human had to be photorealistic and had to work in a complex environment.
The fact that this had to be an action sequence meant that the sequence had to be rendered in real-time with low latency for continuous feedback and a virtual camera.
The final output was supposed to be going for a theatrical release and hence, the output had to be coming in a photorealistic format to look absolutely real.
To create the project in a very cohesive way, the path for the bike was created in which the navigation would be taking place.
When it came to the idea of digital Humans, Viga ET had the inbuilt scanning stage where we were able to scan the actor Mr Sudeep to give an approximate mesh that acted as a guide, and then take the mesh and work in collaboration with headhoppers Studios where the entirely created digital human character was available. The digital human character was later processed with clothing, hair, and materials to make it look realistic under specific lighting conditions.
The bike assets were scanned from the real world with photogrammetry. And the photogrammetry scans helped bring the whole bike asset to an accurate scale model. The actual scale model was then textured and optimized to replay in real-time. The character and the bike were later keyframe animated in a continuous feedback loop.
Overall, the project of Vikrant Rona required immense computer resources at the final render stages. Mainly because of creating 4K renders with very high sample rates. The team at GPUhub were extremely helpful in the scaling up of resources.