Category: Project 2

For this work, this is my first attempt to use AI and 3D scanning to assist in video production, and it is also my first attempt to combine animation and real footage. This is an experimental short film that will open up a different path for my FMP. The tone of the whole film is lost and helpless, but it also reflects my true emotions as a director very well, and this short film can resonate with me very well. However, as a short film that brings in the director’s personal emotions, I think it is more difficult to resonate with the audience, especially when adding real footage, which will make the audience have different emotions, which may lead to mixed reviews of the short film.

In fact, when I was making the last shot, I made two different endings. The first one is as shown in the finished film, the protagonist leaves the camera and focuses on the leftover apple. But the other ending is that when the protagonist leaves, the protagonist takes the leftover apple. The meanings of these two endings are completely different. The first one is that the protagonist has not come out of the mental torture. Although the whole scene has become warm, the apple that symbolizes the protagonist’s rotten heart is still there. This also means that the protagonist only temporarily calmed his mental state with some self-comfort, but the root problem was not solved. When the protagonist took the apple, it meant that the protagonist solved his mental trauma. I used the first shot after thinking for a long time. This is probably also a presentation of my own spirit. I think it is difficult for a patient with mental illness not to rely on external forces to complete self-salvation.

This is another end:

In this part, I render all the existing materials and do the final editing and post-processing in Premiere Pro.

In the first and last shots, I tried to use different lens focuses. The first shot focused on the apple in the foreground, which made the whole picture full of uncertainty. In the last shot, I focused on the character in the back to focus on the emotional changes of the character.

I then removed the color from the real footage, darkened the brightness, and added some vintage film effects.

Regarding the soundtrack, I asked my good friend to help. We discussed the soundtrack for a long time. We determined the climax of the whole film. I explained to my friend the differences in time, space, and vision of all the shots.

I want to add the character’s background story to this short film, but it is difficult to explain the character’s background story within the limited time, so I plan to use photos as an object. My idea is to use photos hanging on the wall to show the protagonist’s past experiences. This will be an empty shot. However, due to time constraints, it is difficult for me to make different clothes for the character and different scenes in different photos, so I used AI.

I searched online and finally identified several AI software for generating pictures: ComfyUI, Leonardo AI, Bing Image Creator, Playground AI. However, in the subsequent production, I found that the common disadvantage of this AI software is that it is difficult to use your own characters to produce AI pictures, and it cannot accurately recognize my original character models.

Then I looked up relevant information and locked on the AI ​​software Astria, which can convert the character images you upload into AI character models, and then use the unique AI character models for AI drawing.

But I still found a serious problem here, that is, if you want to add other characters besides the protagonist to this picture, then the other characters will also become the appearance of the protagonist. I will solve this problem in the future.

In this part, I started to animate my character. First, I used Advanced Skeleton to bind the character, then drew the weights. Then I animated the character in Maya.

After I finished the animation, I exported the animation to abc format, entered Marvelous Designer, and simulated the cloth. Then I exported the simulated clothes to abc format, entered MAYA, and synthesized the entire animation.

In this part, I started to prepare the footage of the reality shots, because the climax of the whole short film is the reality shots, and these reality shots show the character’s past and fantasy. So I need the reality shots I shot to provide enough emotional value to support the whole short film.

I first chose the beach, ​​but the whole shot was a wide angle, and the camera was rotated during the shooting. Although the beach is a good scene for expressing emotions, this shot scheduling did not have the same effect, and even had the opposite effect, so I finally gave up on this shot.

I needed a lens that could carry the character’s emotions in terms of lens language, scene arrangement, and ambient sound, so I thought of the subway. The subway can perfectly cover these needs. I tried to shoot a segment while riding the subway. I chose a 35mm focal length lens setting, looking straight ahead from the perspective of the human eye. The 35mm focal length is generally what the real human eye sees, so this can also reflect the real memories of the character himself.

I then shot two more clips. One was shot facing the window of my house. I chose the afternoon when the sun was shining into the room and there was a breeze blowing, which would cause the curtains to flutter gently. The second clip was a river. I faced the sun and let the sun reflect through the water into my lens. Both clips were shot with a 35mm focal length. The first clip had almost no background sound, while the second clip had a strong wind sound.

After completing the preliminary design of the characters and scenes, I need to modify some details.

Because my short film is a display and simulation of people’s mental state, I need some objects to represent something. I plan to use a bitten apple to represent the character’s mental trauma, and this apple will appear in the foreground of some shots, so the modeling accuracy needs to be relatively high. But in the process of looking for materials, I found that it was difficult to find a model that could meet my needs. I needed to make this apple a little rotten and bitten, which also represents that the character is gradually escaping into the abyss after being traumatized.

So I used Luma 3D for 3D scanning and modeling. First, I bought a bag of apples, took a bite of an apple, and then placed the apple on the table for two days and two nights. I used Luma 3D to scan the apple and exported it to Maya in OBJ format.

Finally, I imported the apple model into my first scene：

I started working on the second scene. My idea for this scene was a huge empty warehouse with a glass window on one of the walls to look out.

There will be many load-bearing columns in the middle of the warehouse, and in the last shot, the character will run forward between two rows of load-bearing columns.

Then I used Marvelous Designer to make the character’s clothes. Because the whole story takes place indoors, I made the clothes very loose, which can also cover up some of the shortcomings and flaws of the character animation during the subsequent cloth simulation.

For this part, I first expanded the UV of the character’s body, and then used substance painter to paint the character’s texture. Because the entire scene scheduling almost never reveals the character’s face, I didn’t spend too much effort on this part. But I want to emphasize that after the first painting, I imported the character into Maya and added hair, and found that the scalp always showed through the hair, so I simply painted the character’s scalp brown.

I then searched online for materials and started to build my scene.

My scene is also very simple, but it requires a large glass window, which paves the way for me to add external lightning flashes and fireworks later.

In this part, I made the protagonist of the whole film. My protagonist was modified from another character of mine. I made a lot of modifications to the face and rewired the body for rigging.

Then, I wanted to try a new way to make the character’s hairstyle.

1. I first made a rectangle.

2. Divided the rectangle into several parts and then exported its UV.

3. Used Photoshop to draw the UV texture.

4. Imported the texture into Maya, then divided and reorganized the rectangle, so that I had a variety of hair strands.

5. Evenly placed the hair on the character’s head and made it fit the character model.

This is the final effect in MAYA：

Entering the preliminary preparations, I first referred to the story that my FMP was going to tell and rewrote the script. This time, the script of Project 2 mainly tells the story of a woman seeking self-salvation in pain, facing the pressure of life and herself, and being overwhelmed. This is also the problem I have faced for a long time. I often compare the present with the past, and then I am full of fantasies about the future. This is a problem that has always troubled me, and it is difficult not to use external factors to improve myself.

I will divide the entire short film into three time points, that is, three acts. The first act is in a room where the protagonist sits on the sofa by himself, experiencing a painful self-struggle. During this period, I will use alternating editing techniques in the room and the protagonist’s fantasy scenes to show the audience. Then in the second act, I will use the material I shot in reality, which represents the protagonist’s past. In the third act, it is still in the protagonist’s room, but the light is much brighter, and the protagonist stands up and leaves.

Although it is a good idea to bring one’s own experience and emotions into the work you create, it also has disadvantages, such as it is difficult to make the audience empathize, which is a double-edged sword.

This is my moodboard. I used the same moodboard from my FMP, because the tone of the two short films is very similar.