Miri Mocap Project

 Posing the mouth:  I must say posing the mouth was way more difficult than posing both the eyes and brows. It was really hard to figure out what control to use for the best results. For example, this pose for frame 124 took me a while to figure out what to do. Had to utilize all controls in the "OH" section to get an optimal shape. The same goes for this pose for frame 523, I thought I could get a decent result using the controls in the "OH" section again, but the lip corner just wouldn't go more upward. And that's how I learned to usage for the Lips push-pull controllers. Dr. Kennedy didn't have the time to talk with every student this week, but I'm confident with the mouth poses, so looking forward to the feedback session next week. Reflection: The mouth-posing process was rather time-consuming for learning what all the controllers would do. Even more time-consuming to learn how to identify which controller to use to construct a good-looking pose....

 Retargeting the brows: The first pass was decent, but I'm not sure why some parts of the retarget for the brows would be raised strangely, I didn't set any keyframes with upward movements and I don't see such behaviors in Analyzer. Frame 36-46 Frame 270-273 Frame 337-355 Frame 417-426 Frame 429-434 Frame 490-502 dropped downwards then rebounded upwards around frame 527-536 I managed to fix the majority of the weird movements, just some jittery here and there which would be easier to smooth out during the cleaning process. Moreover, I left out the brows drop near the end, as the pose was consistent from frame 487 till the end so I can keep that pose stationary during clean up. Retargeting the eyes: The first pass was bad, like terribly bad. For some reason, the eyes were really problematic compared to the brows. Many random movements that weren't in the analyzer, the eyelids should be easy to fix but I might leave out the eyeball and keyframe them manually later in the ...

 Retargeter - brows & eyes posing: The task for this week was rather casual, just simply posing the brows and eyes for the Faceware Retargeter. I did want to see how they would move using the tracking data but not really sure if there's any additional setup with Dr. Kennedy's process. So better leave it for next week. Dr. Kennedy pointed out that the brow position between the neutral frame and frame 1 should be similar and mine was raised too much. This will affect the program calculation when doing the retargeting process. He explained that it could be the actress's head looking downwards that resulted in more movements shown in the footage than there really was in the expression. Another problem he pointed out was to be careful about how much the pupil was actually showing. For example in the image above, both lower eyelids should be raised a bit upwards to better resemble the pupil shape in the reference footage. The mindset I had when working on posing the eyes woul...

 Mouth Tracking: The mouth tracking was very time-consuming, at least for my assigned performance intervals. Miri the character was REALLY expressive with her mouth expressions. I needed way more key poses to define the movements. The first pass was okay but needed more supportive keyframes to maintain the shapes. For example, the upper lip tracker lost track of the shape because it moved in a short period of frames, hence adding additional keyframes to adjust the timing.  The second pass was still not great in my opinion. Another annoying factor that's messing with the tracking was the inconsistency of the tracker position, especially for the lower inner lip. Despite going back and forth between keyframes to ensure the tracker position was as similar as possible, some movements like biting lip followed by opening mouth will result in the tracker losing a consistent position.  Moreover, the change in lighting and shadows just makes it even harder to get consistent. Reflec...