Filtsy Labs is a team of engineers and developers diving deep into the world of AR rendering engine, 3D video and neural rendering.
We love our principle of crafting from scratch, blending the excitement of pure coding with a responsive, customizable AR engine, creating seamless virtual experiences.
Modern video codecs use an 8-bit per pixel per channel capacity to store information. Used to be great until depth-mapped videos became a thing. Now, that 8-bit space for depth channel only capable to map objects bigger than 3.9 cm. As a result you get a low detailed depth map.
We figured out how to record a more detailed depth map while preserving video compression efficiency.
The result is our video codec utilizing a 16-bit capacity for depth channel, so we can now capture depth maps with details as tiny as 0.15 mm with the same compression level!