Advanced Animation

 Advanced Animation 

2026 

Lee Jong Yun / 0358356 

Advanced Animation

Exercise 1: Bouncing Ball Animation

For this exercise, I animated four different balls — soccer, ping pong, bowling, and beach ball — using the provided multiball rig in Blender, with the goal of showing how weight, elasticity, and size change the way an object moves through space. Before touching the rig, I studied reference footage for each ball to understand their real physical behavior: how a bowling ball barely deforms and loses very little energy on impact, how a ping pong ball bounces fast and erratic due to its low mass, how a soccer ball has a mid-weight bounce with decent hangtime, and how a beach ball floats and hangs in the air far longer than its size would suggest because of air resistance.

Working through this exercise taught me a lot about the core fundamentals of animation — timing, spacing, and arcs — and how they aren't just abstract principles but direct translations of physics into motion. I learned that timing (the number of frames between poses) controls how heavy or light an object feels, and that spacing — how close or far apart the in-between frames are — is what actually sells the slow in and slow out on each bounce. Animating four balls with very different physical properties side by side really helped me see how the same rig can read as completely different objects just by changing timing and spacing choices.

I also got more comfortable with the render and composite pipeline — rendering PNG sequences with alpha channels out of Blender, then bringing them into After Effects for compositing and Premiere for final editing with labels and transitions. That workflow was new to me in this context, and doing it four times over reinforced it.

Where I'm weak, and what I want to upgrade:
My main weak point is the motion path of the ball — the arcs aren't as smooth or as physically convincing as they could be. In a few of the bounces, the trajectory looks slightly angular or inconsistent rather than following a clean, continuous parabolic curve, which breaks the illusion of realistic motion. This likely comes from not refining the Graph Editor curves enough after the initial keyframing — I mostly relied on the rig's default interpolation rather than manually adjusting the motion path curves to get true smooth arcs.

To upgrade this, my plan is to go back into Blender's Graph Editor and Motion Path visualization tool to check each ball's trajectory directly in the viewport, and clean up any dips or kinks in the curve. 

---

Exercise 2: Ball with Tail Animation

For this exercise, I animated a squashed-and-stretched ball with a tail using the provided rig, with the focus shifting from pure timing/spacing.

Before animating, I studied the reference video to understand how the tail should lag behind the ball's direction changes, building up drag as the ball moves and then whipping through with overlapping action once the ball stops or changes direction. I planned the key poses on paper first, thinking through where the ball would show anticipation before a movement, and how the tail's flexibility would need to be staged across multiple joints rather than moving as one rigid piece.

Through this exercise I learned how much overlapping action and follow-through contribute to making something feel alive rather than mechanical.

animating and timeline

using motion path to smooth the movements and time the actions

Where I'm weak, and what I want to upgrade:
My weak point is bringing the ball and tail to life through the motion path, right now the movement doesn't fully read as organic, and I want to push the animation to feel more advanced and believable. Specifically, I think the issue is that my tail's motion path isn't building up enough of a smooth.

To upgrade this, I plan to:

• Use Blender's Motion Path tool on each joint of the tail (not just the ball) to visually check that each segment's curve is offset and staggered correctly down the chain, rather than moving in unison.
• Push the timing offsets further between joints so the delay reads more clearly, and refine the Graph Editor curves so the whip and settle happen with more distinct acceleration and deceleration, rather than uniform speed.
• Exaggerate the drag slightly more on fast direction changes, since subtlety at this stage tends to read as stiffness rather than restraint.
• Pay closer attention to how each tail segment overshoots and settles independently, so the follow-through cascades down the tail instead of resolving all at once.

---

Exercise 3: Character Pose Animation (Happy, Sad, Angry, Scared)

For this exercise, I created four emotional poses — Happy, Sad, Angry, and Scared — using a rigged character in Blender. Unlike earlier exercises focused on motion, this one was about appeal and storytelling through a single pose, so I focused on clear silhouettes, strong lines of action, and exaggeration to make each emotion read clearly.

I learned that a good pose needs more than an accurate expression — it needs exaggeration and secondary action to really sell the emotion, and that avoiding symmetry (twinning) is key to making a pose feel alive rather than stiff.

Happy pose-

reference image-

Sad pose-

reference image-

Angry pose-

reference image-

Scared pose-

reference image-

Where I'm weak, and what I want to upgrade:
My weak points were facial expressions and finger movements. My poses relied mainly on body language, while the face and hands stayed too neutral and generic. To upgrade, I plan to study expression references more closely (eyebrows, eyelids, mouth shape) and pose fingers deliberately to match each emotion, like clenched fists for Angry or curled fingers for Scared.

---

Project 1: Walk Cycle Animation

For this project, I animated a vanilla walk and an attitude walk using the Snow character rig in Blender. I referenced walk cycle videos to identify the key extreme poses — contact, down, pass, and up — and used pose-to-pose animation for the main walk action, with straight-ahead animation to add flexibility and secondary movement.

work process

Outcome -

Where I'm weak, and what I want to upgrade:

I want to upgrade the visual presentation by adding camera movement instead of a static shot. A moving camera (like a subtle tracking or dolly shot following the character) would add more production value and make the walk cycles feel more like a finished piece rather than a rig test. To do this, I plan to animate a camera path in Blender that follows the character's forward motion, keeping the framing consistent while adding a bit of dynamism through slight parallax or a tracking angle.

---

Project 2: Cartoon Jump Animation

For this project, I animated a cartoon jump using the Snow character rig in Blender, keyframing the key extreme poses — neutral, anticipation, push up, jump (air), contact/landing, overshoot, and recovery. The focus was on exaggeration and clarity, pushing the line of action and silhouette at each pose to sell a cartoony feel rather than a realistic one, along with applying variation in timing (three speeds) and moving holds to add clarity to the action.

This project taught me how much exaggeration and clear extreme poses matter for cartoony motion — a jump reads as "cartoony" not just from how high the character goes, but from how much the poses stretch, squash, and anticipate before and after the peak.

Where I'm weak, and what I want to upgrade:
My jump currently looks a bit weird/off, and I've traced the issue back to my blocking pass — the extreme poses themselves aren't solid enough yet, so no amount of refining afterward has fixed the motion. To upgrade this, I plan to go back to blocking and rebuild the poses one at a time: making sure anticipation, push up, air pose, and landing are each strong and clear on their own (checking each pose in isolation, not just in motion), before touching the timing or moving on to spline/polish. Fixing the foundation at blocking first should resolve the "weird" feeling once I move back into refining arcs and timing on top of it.

---

Final Project: Physical Action Sequence Animation

For my final project, I animated a physical action sequence in Blender featuring martial arts-style movement, following the full pipeline: storyboarding, blocking with constant interpolation, polishing/splining, and finally camera staging with a mix of full, medium, and close-up shots. This project brought together everything from the earlier exercises — timing, spacing, weight, exaggeration — into one continuous, staged sequence rather than isolated actions.

Working through the three progressions (Blocking, Polishing, Final Animation) taught me the value of separating the animation process into distinct stages instead of trying to get everything right at once. Blocking with constant interpolation forced me to commit to strong extreme and breakdown poses first, and only then move into refining the curves and adding flexibility in the polishing pass. I also learned a lot about staging with the camera — choosing shot types and angles that support the action rather than just capturing it from a flat, static viewpoint.

Where I'm weak, and what I want to upgrade:
My main weakness showed up in the flipping/rotation movements for the martial arts stunts. I animated by moving the rig visually without thinking about the actual rotation math behind it — for example, going from a -500 degree rotation to a -90 degree rotation in just 10 frames, which made the flip look physically wrong and jarring instead of fluid. I realize now I was eyeballing the rotation instead of calculating it properly.

To upgrade this, I plan to:

• Study rotation degrees and angular velocity so I understand how many degrees a body should realistically rotate per frame during a flip, based on the timing and speed of the action.
• Manually check and adjust rotation values in the Graph Editor rather than relying on default interpolation, making sure the degree change between keyframes is consistent and physically sensible.
• Polish the motion afterward so it reads as fluid and human rather than robotic, paying attention to acceleration and deceleration through the rotation, not just a constant speed spin.

---

Module Reflection

This module took me through the full animation pipeline, from simple physics-based objects to a complete character-driven action sequence. Starting with the bouncing ball, I built a foundation in timing, spacing, and arcs. The ball with tail pushed me into drag, overlap, and follow-through. Character posing taught me appeal and storytelling through silhouette and exaggeration. The walk cycle and cartoon jump projects built on that with weight, attitude, and cartoony exaggeration. The final project brought everything together into a staged action sequence with full blocking, polishing, and camera work.

What I've learned:
Across these exercises, I learned that believable animation isn't just about moving something from A to B — it's about applying the right principles at the right stage. Blocking needs to be solid before polishing, poses need to be strong before adding motion, and every part of a rig (tail, fingers, face, limbs) needs its own attention rather than being left as an afterthought. I also learned to think more technically, not just visually — understanding things like rotation degrees and motion paths, not just eyeballing the movement.

What I should do in the future:

• Always study reference and plan poses on paper before jumping into Blender.
• Fix blocking issues at the blocking stage, not later in polish.
• Check motion paths and rotation values directly, not just how it looks in the viewport.
• Give more attention to smaller details — fingers, facial expressions, secondary action — since these sell the emotion and realism as much as the main pose.
• Treat camera work as part of the storytelling, not just a way to capture the animation.

Overall, this module pushed me from understanding animation principles individually to applying them together in a full sequence. My main growth area going forward is precision — being more deliberate and technical in how I build and refine motion, rather than relying on instinct alone.