![[Experiment] Prompts to Control Areola Size | Thorough Verification with 63 Images](/tips/areola-size-control-test/b03_huge_areola_s1_0_0000_2053695854357871006.webp)
Conclusions First
Areolae can be made larger but not smaller. This mirrors the asymmetry seen in the areola color experiment — “darkening > lightening” — now confirmed for size as well.
Effective Prompts (enlarging direction)
Ineffective Prompts (shrinking direction)
small areola, tiny areola, petite areola, small nipples, coin-sized areola — none showed a clear difference from the control. Only tiny areola showed a slight shrinking tendency in 2 out of 3.
Experiment Design
| Parameter | Value |
|---|---|
| Model | z-image-turbo (6B, photorealistic distillation) |
| Steps | 8 |
| CFG | 1.0 |
| Size | 1024x1024 |
| Seeds | 3 fixed (shared across all conditions) |
| Per condition | 3 images |
| Total | 21 conditions × 3 seeds = 63 images |
Base Prompt
A. Control
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Areola size is medium (standard area ratio relative to breasts). Circular. This is the baseline.
B. Size Adjectives (Enlarging Direction)
b01: large areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
2 out of 3 slightly larger. seed 3 shows minimal difference.
b02: big areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
All 3 slightly larger. More stable than large. Side effects also minimal.
b03: huge areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
All 3 slightly to noticeably larger. seed 2 shifted to a chest close-up composition, making the area increase prominent.
b04: wide areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
All 3 slightly larger. Some horizontal spread. Similar to large.
b05: oversized areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
All 3 slightly to noticeably larger. The most consistently large in Group B. However, “oversized” can trigger clothing/fabric props as a side effect (same tendency as observed in the breast size oversized experiment).
Group B Summary
| Condition | Areola Size | Stability | Side Effects |
|---|---|---|---|
| large areola | Slightly larger | 2/3 | Small |
| big areola | Slightly larger | 3/3 | Small |
| huge areola | Slightly–noticeably larger | 3/3 | Composition change |
| wide areola | Slightly larger | 3/3 | Small |
| oversized areola | Slightly–noticeably larger | 3/3 | Clothing triggering |
big areola offers the best balance of effect and side effects for practical use. For maximum size, use huge or oversized.
C. Size Adjectives (Shrinking Direction)
c01: small areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Equivalent to control. No effect.
c02: tiny areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
2 out of 3 (seed 1, 3) showed slight shrinking. The only weak effect in Group C.
c03: small nipples
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Equivalent to control. “nipples” refers to the nipple, not the areola, so areola size is unaffected.
c04: petite areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Equivalent to control. No effect.
Group C Summary
Shrinking is nearly impossible. Only tiny areola shows a weak effect.
D. Shape Descriptors
d01: puffy areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Medium size. seed 1 shows a slight raised appearance. Shape change is minimal.
d02: puffy nipples
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Nipples tend to protrude slightly, but no effect on areola size.
d03: flat areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Slightly small to medium. Tends to render a flatter surface, with more shrinking effect than Group C.
d04: raised areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Medium to slightly larger. Some raised appearance, but limited as size control.
d05: cone-shaped areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
No effect. The cone-shaped instruction was not reflected.
Group D Summary
Shape descriptors have nearly no effect on size control. puffy areola shows a weak raised effect, and flat areola shows a weak shrinking effect.
E. Stacking / Emphasis
e01: very large areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Medium to slightly larger. Adding “very” shows no clear additional effect over large.
e02: extremely wide areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
All 3 slightly to noticeably larger. Consistent enlargement across all 3. The most stable enlargement effect in this experiment.
Note:
extremely wideshowed a consistent enlarging effect in all 3 images, outperformingbigandhuge, which was unexpected. Speculatively, the “horizontally wide” nuance ofwidemay be read as an area increase.
e03: oversized huge areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Similar or greater than e02. Breast size itself also slightly larger. Stacking effect is observable, but the clear difference from e02 is small.
F. Specific Size Descriptions
f01: areola covering half the breast
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Slightly to noticeably larger with enlargement effect. However, all 3 shifted to chest close-up composition, and seed 3 showed granular artifacts. Token efficiency is poor (5 tokens).
f02: silver dollar areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Equivalent to control. seed 3 showed coin-patterned background art. “silver dollar” did not convey size information.
f03: coin-sized areola
| seed 1 | seed 2 | seed 3 |
|---|---|---|
 |  |  |
Slight shrinking tendency but inconsistent.
All Conditions Summary
| Condition | Size Change | Stability | Recommendation |
|---|---|---|---|
| (Control) | Medium | — | — |
| large areola | Slightly larger | 2/3 | △ |
| big areola | Slightly larger | 3/3 | ◯ |
| huge areola | Slightly–noticeably larger | 3/3 | ◯ |
| wide areola | Slightly larger | 3/3 | △ |
| oversized areola | Slightly–noticeably larger | 3/3 | △ Side effects |
| small areola | Medium (no change) | — | × |
| tiny areola | Slightly smaller (weak) | 2/3 | △ |
| small nipples | Medium (no change) | — | × |
| petite areola | Medium (no change) | — | × |
| puffy areola | Medium (weak raised) | — | × |
| puffy nipples | Medium (no change) | — | × |
| flat areola | Slightly smaller (weak) | — | △ |
| raised areola | Medium (weak raised) | — | × |
| cone-shaped | Medium (no change) | — | × |
| very large | Medium–slightly larger | Unstable | × |
| extremely wide | Slightly–noticeably larger | 3/3 | ◎ Best |
| oversized huge | Slightly–noticeably larger | 3/3 | ◯ |
| half the breast | Slightly–noticeably larger | 3/3 | △ Side effects |
| silver dollar | Medium (no change) | — | × |
| coin-sized | Slightly smaller (weak) | Unstable | × |
Note: Enlarging works but shrinking is nearly impossible — the same asymmetry seen in the areola color test. This consistent bias toward the “increase / darken” direction suggests that shrinking likely requires a LoRA or a different approach.
Practical Recommendations
Larger Areolae
Smaller Areolae
Unfortunately no effective prompts were found. tiny areola shows only a weak effect.
Size Control Asymmetry
The asymmetry confirmed in the areola color experiment is replicated here for size:
| Direction | Best Prompt | Effect |
|---|---|---|
| Larger | extremely wide areola | ★★★ |
| Smaller | tiny areola | ★☆☆ |
| Darker | chocolate areola | ★★★ |
| Lighter | peach areola | ★☆☆ |
The model is strong in the “add/darken” direction and weak in the “reduce/lighten” direction.
![[Verified] Image Generation Prompt Best Practices](/tips/prompt-best-practices/cover_0_0000_4517457392071889496.webp)
![[Experiment] Prompts for 'Dark Areola on Fair Skin' | Thorough Verification with 54 Images](/tips/areola-color-control-test/d04_fair_black_s1_0_0000_2053695854357871006.webp)
