Clean Communication of Motion Intent in Live2D

Direction of a pose

The trajectory between the starting pose and the destination pose needs to look appealing.

This depends on:

1. The motion range built during rigging.
2. Timing and spacing when keyframing animation.

When transitioning from pose to pose, a directional force is created. This force informs the following drag, stretch, squash, and secondary deformation, naturally conveying a sense of direction.

Forward Lean in POS Z

One of the most common simple trajectories used in VTuber models is leaning forward.

Some rigs achieve this by bending at the waist, while others push the body forward from above the knees.

Since VTuber models mainly rely on simple motion capture and simulated camera depth, clear communication of motion depends heavily on the performer's acting. The principles are the same as animation.

VTubers who want to improve their performance should study animation acting.

A forward lean should always have a clear intention behind it.

For creators with less reliable tracking, it can be helpful to add a dedicated forward-lean hotkey to communicate intent more clearly. This kind of hotkey can also be used for scene transitions and other presentation tricks during streams.

Rigging vs Performance

If the fundamental rigging is already solid, whether a model is used well ultimately depends on the performer.

Even a perfect model may only perform at 50–70% of its potential if the user lacks performance skills.

For VTubers who do not learn how to tune their tracking setup or improve their acting, even an excellent model may only reach 20–30% of its potential, which is a bit unfortunate.

Animation Cleanup

Keyframed animation can often communicate ideas more cleanly.

For simple motion and expressions, VTuber performances can be recorded in VTube Studio and then brought back into animation software for cleanup.

For dedicated performing animation sequences, motion capture is often unnecessary, although some productions combine motion capture and hand animation depending on the requirements.

Clarity of Direction

Motion without purpose feels awkward.

When an action lacks intention, it either feels forced or leaves the audience unsure what the movement is supposed to communicate.

This is often seen in model showcase videos where a character remains in a static pose while exaggerated physics are added everywhere to demonstrate "mobility."

That said, things like floating jellyfish-style arm physics are sometimes still expected by clients. If the goal is simply to avoid stiffness, a small amount of movement is usually enough.

For VTuber models, effort is often better spent on perspective simulation and material-like deformation.

Overall, motion should exist to serve communication.

Hopefully people's standards and visual literacy will continue to improve over time.

Planning Motion as a Whole

Anticipation

Before the main action begins, there should be anticipation based on the purpose of the movement.

When keyframing:

• Add a preparation phase.
• Hold anticipation poses slightly longer.
• Use compression in the opposite direction before release.

During the release:

• Hold the midpoint slightly longer.
• Create a readable arc that the eye can follow.

For a smooth arc from start to finish, at least five key poses are usually needed.

Apply ease-in at the beginning and ease-out before the ending pose.

Finish with a small overshoot in the opposite direction, often through perspective rotation or volume deformation.

Blocking Key Poses

Key poses should be planned according to the storyboard.

The animator determines:

• When transitions occur.
• How long poses are held.
• The rhythm and pacing of the performance.

These decisions create the unique cadence of the animation.

Idle Breathing

A character does not need constant special actions.

But a character should not appear incapable of breathing.

Even idle animation should communicate small pieces of information about the character's personality and traits.

Leading and Following Motion

By determining what moves first and what follows after, it becomes easier to decide where to switch between IK and FK.

For standing VTuber models, leg IK alone is usually sufficient because the feet rarely leave the ground.

However, for more complex actions such as:

• Jumping
• Standing ↔ Sitting
• Walk → Run → Jump

FK and IK often need to be switched back and forth.

Because Live2D is fundamentally a 2D software, these transitions usually require duplicated parts, separate parameters, and separate bone setups to support both systems.