The film uses a total of 1,400 CG special effects lenses, costing nearly 200 million US dollars, so many CG lenses, fully prepared for the 3D version. In the original version, these CG special effects shots often retain every part of the scene, scene engineering files from multiple camera angles, and three-dimensional depth information. Etc.) have established three-dimensional models, and then break these models. The 3D special effects team rearranged these shots and used the previously built 3D scenes to make the camera's angle, angle of view, and focus point accurately and realistically, and re-rendered and output high-quality pictures. Debug and preview, and strive for the most shocking, real, and most in line with the physical space relationship.
At the beginning of the film, a large number of photos were collected from the crowd, and then CG models were established for them. For the main characters, based on the scan data from Eyetronics, a powerful database and model library were established, and new software environments and algorithms were developed, including GPUs. The special hardware environment, including processing matrix, Solid State Drive, high-speed internal network, etc., enables the production of the film to achieve a double increase in efficiency and quality.
In addition to the CG special effects lens, the original shooting retains a large amount of 3D shooting material, which will also be used in this 3D production. A large number of depth information maps have been extracted during the original shooting. The conversion directly guarantees the high reproducibility and authenticity of its conversion quality, and deeply repairs the previous poor effect due to the immature technology, which greatly increases the 3D effect.
