1. Painting the face for the two characters

2. Sculpt out lip position for different sounds. Including the vowels in Japanese (a,e,u,e,o) and consonant position for lip closing

3. Sculpt out some expressions

4. insert into Maya and add their hair

5. adding riggings to the model for doing movements

6. Adjust different lip shape according to the audio, render out the animation and input into after effect

7. Add background, foggy effect and sound to the video and render out. Then input into Premiere to add opening and ending

This is the final video and the presentation






This 3D animation takes place inside the bathroom of a nightclub, where a girl is high on MDMA and sees herself in the mirror. Her reflection in the mirror shows her jaw moving, eyes rolling, and slowing dancing to the music until she hears a girl screaming, which pulls her out of that trance moment and walks out of the scene. It is not know whether she has been drugged or decided to take the drug herself, but this is not the matter of importance. This scene is a reflection with many questions opened for the audience to resolve. I want to express the mystic experience of drugs which can be both scary and seductive.

This project was 3D modeled in Maya and Mudbox, and the end sequence was compiled and edited in After Effects.

After fixing and re-texturing my 3D model, I finally started to create the expressions for my animation. Here you can see the final version of my model. I also added painted make-up on its texture for the club scene.

This are some of the expressions that I sculpted in Mudbox:


After modeling the expressions in Mudbox, I sent my model to Maya, where I continued by rigging the neck and head and animating it with the facial expressions. As to the lighting, I decided to add one spotlight and one pointlight to my scene. When it came to rendering, I rendered the view from three different cameras. One for the front (the mirror reflection), one for the back view, and the last one for a close up of the mouth and eyes. Below, you can see some of the camera shots:



In After Effects, I added several different effects to make the scene more psychedelic in a sense and to finalize. Some of these effects include Echo, Gaussian blur, Tint, Opacity Flash, etc. I also used masks to create the broken mirror illusion. For the setting, I added some images and textures such as the bathroom image, a texture of stains on glass, broken glass, and a reflection to make the mirror more realistic.

Some of the audio for this piece was downloaded from Freesound.org and edited in Audacity. The main track of this piece is Sun My Sweet Sun by Konstantin Sibold. I also added an audio sample of people talking, of a creaky door, of a door closing, and of a girl screaming to create a more realistic ambiance of the scene. I added a Low Pass Filter to my main audio track in order to create the sensation as though the sound was coming from outside of the bathroom in the club.

These are some shots of my sequence.

Overall, this was probably one of my favorite projects I have done. I feel like I have learn a lot doing this project and would love to continue it further in the future.

DSFA | Assignment 4 (Final Project) | Me, Myself and I | Presentation (Yumin Cai)

Here’s the link for the presentation:


I chose to do the lip sync as the main content for my final project and the inspiration is one of my favorite scientific movie series Tron. I chose several lines of the heroine Quorra in ending part of the movie Tron: Legacy. More details could be seen in my presentation file above.

Here is the preview of it:

Here is the video:

DSFA | Week 14 | Assignment 4

For this project, I decided to do a lipsync from Disney’s Emperor’s New Groove. I chose the scene where Yzma is checking that Kronk has the poison to kill Emperor Kuzco. The short section I modeled my sculpture on is when Kronk does a little “monologue” about which poison he has. The script follows as:

“Oh right, the poison. The poison for Kuzco. The poison chosen specially to kill Kuzco. Kuzco’s poison. That poison? … Gotcha covered.”

For the sculpt layers, I made fourteen different layers:

  1. His confused starting look
  2. His gasp when he realizes what Yzma is asking
  3. Oh
  4. R
  5. I (sounds like eye)
  6. T | S | N | E (sounds like ee) *
  7. Th | L *
  8. Uh | ih *
  9. P
  10. Oi
  11. F *
  12. OO
  13. Ch | Sh *
  14. Ah (sounds like apple)

The layers with an asterisk (*) are layers where one mouth position works for several different sounds, as shown with the different letters or sounds in the “title” of the layer. With the F layer, it also works for the V in “covered”. It is not named “F | V” because I was originally unsure if I was going to use that part of the clip.

After sculpting all the layers, I realized I had sculpted all the eyebrows to very beginning expressions that don’t stay throughout the piece. In order to complete my project on time, I had to make the choice to keep them the way they were.

I exported my Mudbox file to Maya. Here, I started using the shape editor to start keying my frames. At first, this was very difficult. I was looking at the reference video to try to key the mouth movements. I also had to key in where the layer would start, be at its max, and where it would end. This way, the layers would blend nicely with each other and look natural. When I first started keying, I forgot to key the first 0 position, so when I played the animation, three layers were laying on top of each other at different strengths making my project look very wrong. I soon found the reason and fixed it.

After meeting with Professor Chen, I was able to import my audio into Maya which made everything so much easier! I didn’t have to keep estimating from the reference video (which I had put into After Effects to see the frames).

In order for me to wrap my head around where things needed to be keyed, I wrote out all 341 frames in my sketchbook and put the layers start, max, and end positions above the corresponding frame.

After finishing all the keying, I noticed that when I tried to do some head movements and rigging (and just reopening my project from day to day) my program was running incredibly slowly. It was found that my file size was way too large. After talking to Professor Chen, I deleted some unnecessary layers that were in my file, optimize the file, and cut its size down by almost half. It made working much easier.

Professor helped me do some exports in After Effects because my rendering (done in two segments) was causing some sort of exporting glitch. I was able to edit the one we exported in class to make it into my final. I added a title slide for about four seconds, then my animation plays. My model also has a background picture of the room where this conversation is taking place.

PRESENTATION: https://docs.google.com/presentation/d/1l9dFiUiwL-J7LF0GhUSyAEXz1ZnUSJsqjmjiyqXA3bo/edit?usp=sharing

DSFA Assignment 3: Augmented me_ Wentao Wang

These are the 14 facial expressions modeled in Mudbox. The overall feelings are fine but it will be better to have some additional details to make the expressions more realistic. For example wrinkles on the side of the mouth in on the smiling face. In addition to that, I haven’t added the teeth models in them so some expressions might seems weird.

angry face


sticking tongue out



ugly smile









The second step is to send the models in Maya and animate with them.

This is the normal face after sending into Maya, with teeth model added onto it.

I added three spotlights to shine light onto different angles towards the face to achieve a better realistic effect.

I also added another camera to work on camera animation. The left part of the screen is the camera view and the right part shows the actual working view from the perception camera.

This is an example of the working view when making animations with different expressions.

This is the final video. There are still factors that need improvement. One factor is the camera moving. I didn’t have much time to make a more up to rhythm animation, so the camera animation might look a bit messy.



DSFA | Assignment 3 | Me, Augmented | Presentation (Yumin Cai)

Here is the link to my presentation in class for assignment 3 and the whole work process so far:

In-class presentation for assignment 3:


The whole work so far:


The movie for the expressions:


In this assignment, we have viewed our own facial expressions and tried to apply them to our own models. It is quite interesting to explore how the muscles, eyes, eyebrows, and mouth move when we are smiling, angry or making funny faces. In addition, as we are required to play with the animation to show the moving process of the facial expressions, I also try to combine the eye and head movements with the facial expression to show some more complex motions.

For example, such as the three steps of showing ”at a loss with a little bit shocked”, which is also one of my favorites:

The original one without facial expression:

The “a little bit shocked” one:

The three steps(including the movement of eyes, head, cameras, and light):

Eyes move up to stare at the front:

Look to the left:


Then look to the right as if seeking for something:

There are also things that need to be improved:  take the lights for example, as a result of experimenting with light from different angles, my shadows take a quite important part of my work. However, the shadow turns out to be pixelated. I’ve searched for the reasons on the internet, and found this source extremely helpful for me to understand the different effects on shadows with different light settings:


To sum up, there are two possible solutions:

  1. Choose Raytraced Shadows to take place the Depth Map Shadows to improve its smoothness.
  2. Increase the resolution or filter size level.

I’m still experimenting with these solutions and will try to fix this problem in the coming final project.

All in all, thanks for professor Chen and my classmates’ help when meeting up with the problem in this assignment. Also, thanks to professor Chen’s advice for my in-class presentation, since lacking confidence has always been a big problem for me, I will try to overcome this shortcoming especially when introducing my project to the public in the future!