Category: Design for Animation, Narrative Stuctures & Film Language

Let’s go back to the 2020 Nvidia launch for a moment, the lifelike announcement is good enough to fake it. The emergent technologies, Omniverse, are used for the announcement of the simulation. This new technology focuses on the animation part and the other part such as modelling is almost the same as before. Thousands of photographs were taken from different directions using a large number of cameras when making the model of the speaker. Same as before workflow, using Maya and Zbrush to make a visual character. Then it’s time for the technical climax, building on the original audio-driven animation technology, NVIDIA has adopted the new Face video to video technology. In simple terms, the technology maps a captured real video onto a virtual model. The realistic real-time textures reduce the uncanny valley. The consequent problem is that the light source of the mapping is fixed so that the highlights do not move correctly with the camera when the view is rotated in 3d software. The Nvidia used AI deep learning techniques. Instead of calculating light changes through a physical renderer, multiple videos were shot in advance for the computer to learn the distribution of highlights at different angles. Although this approach cannot calculate real-time light but for global lighting, getting results that look realistic is what we are ultimately aiming for. A similar method is used for the capture of body action. The speakers were asked to perform several times and videos of the performers’ previous presentations were used to learn the performers’ movement habits. After learning by the computer, once the speaker’s voice has been captured, the action can be selected from the corresponding actions. Each syllable corresponds to a number of different actions, and after computer calculation, different actions are selected to correspond to different scenes. Once Nvidia has enough character action models, it will be possible to directly use actions that match the character’s personality in the future in the film, television or gaming industries. This workflow will significantly reduce the product development cycle.

There are of course limitations to such a technology, as there are scenarios where humans use a particular language of movement to express emotions, which the technology is currently unable to do. And the light mapping by deep learning is very limited, if there are other special light sources in the scene, the computer will get incorrect results, because the result is not obtained by real light calculation. But I believe the cutting edge technology of the future will better help artists create their work.

I will start my analysis with a relatively simple film in the war genre. The first movie is the Three Kingdoms: Resurrection of the Dragon (2008) which is released in China and another one is Flags of Our Fathers (2006) which is released in America. Both two films describe a fight between two countries.

The first film follows the legendary life of Zhao Yun, the famous general of the Shu Han Dynasty during the Three Kingdoms period in ancient China. In his early sixties, Zhao Yun, a famous general of the Shu Han Dynasty, returns to the mountains in order to help Shu recover the Central Plains and unify the country, only to be surrounded by the Wei army on Mount Fengming and to face the last battle of his life. Faced with internal and external problems, even Zhao Yun’s resourcefulness could not save him from defeat at Wei.

The Father’s Flag film depicts the background of the acclaimed photographic work ‘Raising the Flag on Iwo Jima’ which represents the Battle of Iwo Jima. It tells the story of the 5th U.S. Navy Division’s attempt to capture the heights of Iwo Jima in 1945, which was stalled until the fifth day when American troops were badly killed and wounded, and the Japanese were forced to retreat to the island’s caves. The whole process was met with fierce Japanese resistance and heavy losses on both sides. Finally, after a month of fighting to the death, the American flag was planted at the highest point on the island.

I’d like to share one of my favourite movies, Iran Man, which is part of the Marvel Cinematic Universe. I was captivated by the ups and downs of the film’s plot, at the same time, I was impressed by Iron Man’s personality. So I will begin by analysing the narrative vision.

The Iron man has a successful narrative structure. At the beginning of the movie, the director shows us an arrogant billionaire, Tony Stark. Through an arms deal, The protagonist’s sloppy characterisation is quickly established. Then the storyline into a long rasing action to make conflict. Although Tony survived the terrorist attack, he was disoriented for a while. The director creates a conflict within the protagonist and also depicts the contradictions of Stark’s industrial weapons. With the help of Dr. Ethan, he eventually becomes Iron Man and logical bring ultimate antagonist. Then, the film reveals all the secrets at the climax, the fight with Obadiah Stane who is a shareholder of Stark Industries, bent on replacing Tony Stark as Chairman of Stark Industries. The movie end at a press conference the next day, Tony announced that I am Iron Man. The short resolution successful lead audiences to imagine the future and so, the Marvel Cinematic Universe is officially open. As I said at the beginning, the film’s plot is tightly paced and has a sense of rhythm.

Then I will talk about some lens detail of the strong terms of references. Such as the picture below:

Both two views organised the space well and the position of the main subjects rule the RULE OF THIRDS. For picture1, appropriate space in the left frame will better extend the protagonist’s eyesight and at the same time enhance the audience’s focal area. For Picture2, The characters’ positions reflect the character relationships and enhance the plot of the story. Tony, who needs help, is in the bottom right corner of the screen and the assistant in the top left corner strikes the balance of the screen.