Breast Motion Keywords in Topless Jump Rope Prompts — Do They Actually Work?

Key Findings

• All 5 breast motion keywords (bouncing breasts, jiggling breasts, breast motion blur, swaying breasts, dynamic breast movement) produced no visible difference from the control
• Topless + jumping rope alone generates stable jump rope poses consistently (0/3 breakdown rate across all conditions)
• Breast depiction remained static in all conditions — no visible bounce, blur, or deformation was observed
• As still images, “motion” is inherently difficult to express through keywords alone

What You’ll Learn

• Stability of topless jump rope prompts
• Whether breast motion keywords produce visible effects
• Limitations of motion expression in still image generation

Experiment Setup

Base Prompt

Only the {KEYWORD} portion changes per condition. A00 (control) omits this part entirely.

Test Conditions

A00: Control (No Additional Keyword)

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 correctly reproduced the jump rope pose — both feet off the ground, both hands holding the rope
• 3/3 showed static breast depiction — no bounce or blur observed
• 3/3 no anatomical breakdown
• 3/3 barefoot (no shoes specified in prompt)

A01: bouncing breasts

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 maintained jump rope pose
• 3/3 no noticeable difference in breast depiction compared to control — no bounce, deformation, or blur
• 3/3 no anatomical breakdown

A02: jiggling breasts

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 maintained jump rope pose
• 3/3 no difference from control in breast depiction
• 3/3 no anatomical breakdown

A03: breast motion blur

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 maintained jump rope pose
• 3/3 no difference from control — no motion blur effect on breasts
• 3/3 no anatomical breakdown
• No motion blur was observed anywhere in the image

A04: swaying breasts

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 maintained jump rope pose
• 3/3 no difference from control in breast depiction
• 3/3 no anatomical breakdown

A05: dynamic breast movement

NSFW - クリックで表示

seed=42	seed=123	seed=789

Observations:

• 3/3 maintained jump rope pose
• 3/3 no difference from control in breast depiction
• 3/3 no anatomical breakdown

Side-by-Side Comparison (seed=42)

NSFW - クリックで表示

Condition	seed=42
A00: Control
A01: bouncing breasts
A02: jiggling breasts
A03: breast motion blur
A04: swaying breasts
A05: dynamic breast movement

Summary

No difference in breast motion expression was observed across any condition. On the positive side, topless × jump rope pose stability proved extremely high — all 18 images across 6 conditions showed zero breakdowns.

Note: Expressing “bounce” in a still image requires shape deformation or blur, which can be close to what models treat as an artifact to avoid. This is speculative, but it may be one reason motion keywords are hard to reproduce. On the positive side, the topless × jump rope composition proved rock-solid stable, with zero breakdowns across all 18 shots.

Analysis

None of the breast motion keywords produced visible effects. Several factors likely contribute:

• Still image limitations: “Bouncing” and “jiggling” are temporal concepts that don’t translate well to single frames. Video generation models may produce different results
• Strong base composition: The topless + jumping rope combination already produces a stable, well-defined composition, potentially leaving little room for additional keyword influence
• Token efficiency: These keywords consume 2-3 tokens with no visible payoff — better spent on scene description or lighting

The topless jump rope composition itself is highly stable and practical. There’s no need to add breast motion keywords — topless, jumping rope alone delivers consistent results.

Note: Tokens spent on keywords with no visible effect are better spent on lighting or background description instead.

Related Articles

[Experiment] Dynamic Pose Comparison — How Far Can Running, Jumping, and Looking Back Go Without Distortion?

[Verified] Image Generation Prompt Best Practices

The Rules of AI Image Generation Prompts | Word Order, Emphasis Syntax, and Negative Prompt Basics