Designs systems for encoding, scoring, and generating choreographic movement using Laban notation, computational geometry, and procedural animation principles.
Resources
1Install
npx skillscat add organvm-iv-taxis/a-i-skills/movement-notation-systems Install via the SkillsCat registry.
SKILL.md
Movement Notation Systems
This skill provides guidance for creating systems that encode, analyze, and generate human movement for choreography, animation, and movement analysis.
Core Competencies
- Movement Notation: Labanotation, Benesh, Motif notation
- Computational Geometry: Skeletal representation, joint angles
- Procedural Animation: Rule-based movement generation
- Effort-Shape Analysis: Laban Movement Analysis (LMA)
- Temporal Structures: Rhythm, phrasing, dynamics
Movement Notation Fundamentals
The Challenge of Movement
Movement is inherently multidimensional:
- 3D spatial paths
- Temporal evolution
- Body part coordination
- Qualitative dynamics (effort)
- Relational context (other bodies, objects, space)
Major Notation Systems
| System | Strengths | Use Cases |
|---|---|---|
| Labanotation | Complete, precise | Archival, reconstruction |
| Benesh | Compact, visual | Ballet, therapy |
| Motif | Abstract, readable | Teaching, analysis |
| Motion Capture | Exact coordinates | Animation, research |
Laban Movement Analysis Framework
Body Component
What body parts are moving:
Body Organization:
├── Core-Distal (center outward)
├── Head-Tail (spinal connection)
├── Upper-Lower (horizontal division)
├── Body-Half (left-right)
└── Cross-Lateral (diagonal connections)
Body Parts Hierarchy:
Center (pelvis)
├── Torso (spine, chest)
│ ├── Head
│ ├── Shoulders
│ └── Arms → Elbows → Hands → Fingers
└── Hips
└── Legs → Knees → Feet → ToesSpace Component
Where the body moves:
class KinesphereModel:
"""The reachable space around the body"""
DIMENSIONS = {
'vertical': {'up', 'down'},
'horizontal': {'left', 'right'},
'sagittal': {'forward', 'backward'}
}
LEVELS = ['low', 'middle', 'high']
# 27 directions in the kinesphere
DIRECTION_SYMBOLS = {
'place_high': (0, 1, 0),
'place_middle': (0, 0, 0),
'place_low': (0, -1, 0),
'forward_high': (0, 1, 1),
'forward_middle': (0, 0, 1),
'forward_low': (0, -1, 1),
# ... all 27 combinations
}
# Spatial scales
SCALES = {
'near': 0.3, # Close to body center
'mid': 0.6, # General reach
'far': 1.0 # Full extension
}Effort Component
How movement is performed (qualitative dynamics):
class EffortFactors:
"""Laban Effort qualities"""
FACTORS = {
'weight': {
'light': {'sensation': 'buoyant', 'value': -1},
'strong': {'sensation': 'powerful', 'value': 1}
},
'time': {
'sustained': {'sensation': 'leisurely', 'value': -1},
'quick': {'sensation': 'urgent', 'value': 1}
},
'space': {
'indirect': {'sensation': 'flexible', 'value': -1},
'direct': {'sensation': 'focused', 'value': 1}
},
'flow': {
'free': {'sensation': 'fluent', 'value': -1},
'bound': {'sensation': 'controlled', 'value': 1}
}
}
# Basic Effort Actions (combinations of weight, time, space)
ACTIONS = {
'punch': {'weight': 'strong', 'time': 'quick', 'space': 'direct'},
'dab': {'weight': 'light', 'time': 'quick', 'space': 'direct'},
'slash': {'weight': 'strong', 'time': 'quick', 'space': 'indirect'},
'flick': {'weight': 'light', 'time': 'quick', 'space': 'indirect'},
'press': {'weight': 'strong', 'time': 'sustained', 'space': 'direct'},
'glide': {'weight': 'light', 'time': 'sustained', 'space': 'direct'},
'wring': {'weight': 'strong', 'time': 'sustained', 'space': 'indirect'},
'float': {'weight': 'light', 'time': 'sustained', 'space': 'indirect'}
}Shape Component
How the body changes form:
class ShapeQualities:
"""Body shape changes"""
MODES = {
'shape_flow': {
'description': 'Internal shaping, self-oriented',
'examples': ['breathing', 'growing/shrinking']
},
'directional': {
'description': 'Bridge to environment',
'subtypes': ['spoke-like', 'arc-like']
},
'carving': {
'description': 'Sculpting 3D space',
'relationship': 'Interacting with environment'
}
}
AFFINITIES = {
'rising': {'effort': 'light', 'direction': 'up'},
'sinking': {'effort': 'strong', 'direction': 'down'},
'spreading': {'effort': 'indirect', 'direction': 'horizontal'},
'enclosing': {'effort': 'direct', 'direction': 'in'},
'advancing': {'effort': 'sustained', 'direction': 'forward'},
'retreating': {'effort': 'quick', 'direction': 'back'}
}Computational Movement Representation
Skeletal Data Structure
class Skeleton:
"""Hierarchical skeletal representation"""
def __init__(self):
self.joints = {
'pelvis': Joint('pelvis', parent=None),
'spine': Joint('spine', parent='pelvis'),
'chest': Joint('chest', parent='spine'),
'neck': Joint('neck', parent='chest'),
'head': Joint('head', parent='neck'),
'l_shoulder': Joint('l_shoulder', parent='chest'),
'l_elbow': Joint('l_elbow', parent='l_shoulder'),
'l_wrist': Joint('l_wrist', parent='l_elbow'),
'r_shoulder': Joint('r_shoulder', parent='chest'),
# ... etc
}
def get_world_position(self, joint_name):
"""Compute global position from local transforms"""
joint = self.joints[joint_name]
position = joint.local_position
current = joint
while current.parent:
parent = self.joints[current.parent]
position = parent.rotation.apply(position) + parent.local_position
current = parent
return position
def compute_joint_angles(self):
"""Extract joint angles for analysis"""
angles = {}
for name, joint in self.joints.items():
if joint.parent:
angles[name] = joint.rotation.as_euler('xyz')
return angles
class Joint:
"""Single joint in skeleton hierarchy"""
def __init__(self, name, parent=None):
self.name = name
self.parent = parent
self.local_position = np.array([0, 0, 0])
self.rotation = Rotation.identity()
self.constraints = {} # Joint limitsMotion Trajectory
class MotionTrajectory:
"""Temporal sequence of poses"""
def __init__(self, fps=30):
self.fps = fps
self.frames = [] # List of Skeleton states
self.annotations = [] # Qualitative markers
def duration(self):
return len(self.frames) / self.fps
def get_velocity(self, joint_name, frame_idx):
"""Compute instantaneous velocity"""
if frame_idx < 1:
return np.zeros(3)
pos_current = self.frames[frame_idx].get_world_position(joint_name)
pos_prev = self.frames[frame_idx - 1].get_world_position(joint_name)
return (pos_current - pos_prev) * self.fps
def extract_effort_features(self, joint_name, window=10):
"""Estimate Laban Effort qualities from motion"""
features = {
'weight': self._compute_acceleration_magnitude(joint_name, window),
'time': self._compute_temporal_change_rate(joint_name, window),
'space': self._compute_path_directness(joint_name, window),
'flow': self._compute_flow_continuity(joint_name, window)
}
return featuresProcedural Movement Generation
Rule-Based Choreography
class ChoreographyGenerator:
"""Generate movement sequences from rules"""
def __init__(self):
self.vocabulary = self._load_movement_vocabulary()
self.grammar = self._load_grammar_rules()
def generate_phrase(self, theme, duration_beats=16):
"""Generate choreographic phrase"""
# Start with motif based on theme
motif = self._select_motif(theme)
# Develop through phrase
phrase = [motif]
current_beats = motif.duration_beats
while current_beats < duration_beats:
# Apply development rules
if random.random() < 0.3:
# Repeat with variation
variation = self._vary_motif(phrase[-1])
phrase.append(variation)
elif random.random() < 0.5:
# Contrast
contrast = self._generate_contrast(phrase[-1])
phrase.append(contrast)
else:
# Transition
transition = self._smooth_transition(phrase[-1])
phrase.append(transition)
current_beats += phrase[-1].duration_beats
return phrase
def _vary_motif(self, motif):
"""Create variation of movement"""
variations = [
self._change_level, # Same movement, different level
self._mirror, # Left-right reversal
self._change_size, # Larger or smaller
self._change_tempo, # Faster or slower
self._change_direction, # Face different direction
self._fragment, # Part of the movement
self._extend, # Add onto the movement
]
return random.choice(variations)(motif)Effort-Driven Animation
class EffortAnimator:
"""Generate movement with specified Effort qualities"""
def animate_action(self, skeleton, target_position, effort_state):
"""Move body part with specified Effort"""
# Time factor affects duration
if effort_state['time'] == 'quick':
duration = 0.3
ease_type = 'exponential'
else: # sustained
duration = 1.2
ease_type = 'sine'
# Weight factor affects acceleration
if effort_state['weight'] == 'strong':
acceleration_curve = self._strong_curve()
else: # light
acceleration_curve = self._light_curve()
# Space factor affects path
if effort_state['space'] == 'direct':
path = self._linear_path(skeleton.current, target_position)
else: # indirect
path = self._curved_path(skeleton.current, target_position)
# Flow factor affects continuity
if effort_state['flow'] == 'bound':
path = self._add_micro_pauses(path)
# free flow is smooth by default
return self._create_animation(
path=path,
duration=duration,
acceleration=acceleration_curve
)Spatial Pattern Generation
class FloorPatternGenerator:
"""Generate spatial pathways"""
def generate_path(self, pattern_type, space_bounds, duration):
"""Generate floor pattern"""
patterns = {
'circular': self._circular_path,
'spiral': self._spiral_path,
'figure_eight': self._figure_eight,
'diagonal_cross': self._diagonal_cross,
'zigzag': self._zigzag_path,
'random_walk': self._random_walk
}
generator = patterns.get(pattern_type, self._random_walk)
return generator(space_bounds, duration)
def _spiral_path(self, bounds, duration, turns=3):
"""Generate spiral floor pattern"""
center = bounds.center
max_radius = min(bounds.width, bounds.height) / 2
points = []
num_points = int(duration * 30) # 30 points per beat
for i in range(num_points):
t = i / num_points
angle = t * turns * 2 * np.pi
radius = max_radius * (1 - t) # Spiral inward
x = center[0] + radius * np.cos(angle)
y = center[1] + radius * np.sin(angle)
points.append((x, y, t * duration))
return pointsNotation Output
Textual Notation Format
class MovementScore:
"""Text-based movement score"""
def to_notation(self, phrase):
"""Convert phrase to readable notation"""
score = []
for movement in phrase:
notation = {
'beat': movement.start_beat,
'duration': movement.duration_beats,
'body': self._notate_body(movement),
'space': self._notate_space(movement),
'effort': self._notate_effort(movement),
'description': movement.description
}
score.append(notation)
return score
def _notate_effort(self, movement):
"""Effort notation symbols"""
symbols = {
('strong', 'quick', 'direct'): '⚡', # Punch
('light', 'sustained', 'indirect'): '☁️', # Float
('strong', 'sustained', 'indirect'): '🌀', # Wring
# ... etc
}
effort_tuple = (
movement.effort['weight'],
movement.effort['time'],
movement.effort['space']
)
return symbols.get(effort_tuple, '○')Visual Score Representation
Time →
| 0 | 1 | 2 | 3 | 4 |
├─────────┼─────────┼─────────┼─────────┼─────────┤
│ ◊ HEAD │ │ ○ │ │ │
│ ═ ARMS │ ═══════│════ │ ═════ │═════ │
│ ║ TORSO │ ║ │ ║║ │ ║ │ ║║║ │
│ ‖ LEGS │ ‖‖ │ ‖ ‖ │ ‖‖ │ ‖ ‖ │
├─────────┼─────────┼─────────┼─────────┼─────────┤
│ Level: │ High │ Mid │ Low │ Mid │
│ Effort: │ ⚡ │ ☁️ │ 🌀 │ ⚡ │
└─────────┴─────────┴─────────┴─────────┴─────────┘Best Practices
Encoding Principles
- Preserve intent: Capture what the choreographer wants, not just positions
- Layer information: Structure > Shape > Dynamics
- Allow interpretation: Don't over-specify unless needed
- Support variation: Enable controlled improvisation
Analysis Guidelines
- Extract both quantitative (positions) and qualitative (effort) features
- Consider cultural and stylistic context
- Validate with practitioners
References
references/labanotation-symbols.md- Symbol reference for Labanotationreferences/lma-glossary.md- Laban Movement Analysis terminologyreferences/motion-capture-formats.md- Common mocap data formats