# MiniMax H3 Ref2VA Prompt Template
## Reference Assignment
`<Picture 1>` is the environment and master scene reference.
`<Picture 2>`, `<Picture 3>` and `<Picture 4>` all reference the same character, **Subject 1**.
* `<Picture 2>` is the primary full-body reference for Subject 1.
* `<Picture 3>` is a frontal facial reference for Subject 1.
* `<Picture 4>` is an additional facial / three-quarter reference for Subject 1.
`<Picture 5>` and `<Picture 6>` both reference the same character, **Subject 2**.
* `<Picture 5>` is the primary full-body reference for Subject 2.
* `<Picture 6>` is a facial reference for Subject 2.
## Reference Usage
Use `<Picture 1>` to establish the environment, spatial layout, background appearance, lighting and overall scene.
Use `<Picture 2>` as the primary reference for Subject 1's body proportions, overall appearance, hair and clothing.
Use `<Picture 3>` and `<Picture 4>` as complementary references for Subject 1's facial identity.
Use `<Picture 5>` as the primary reference for Subject 2's body proportions, overall appearance, hair and clothing.
Use `<Picture 6>` as a complementary reference for Subject 2's facial identity.
All references assigned to the same subject depict the **same person** and must be combined into one consistent identity.
Character-reference backgrounds are not part of the target environment.
The environment is established by `<Picture 1>`.
## Shot
**Shot 1 — 0.00s–10.00s**
[Describe the initial composition.]
Subject 1 is [position and orientation].
Subject 2 is [position and orientation].
The environment corresponds to `<Picture 1>`.
[Describe the principal action clearly and chronologically.]
Subject 1 [action].
Subject 2 [reaction/action].
[Describe subsequent movement or interaction.]
The subjects maintain their respective identities throughout the shot.
## Camera
[Describe the initial framing.]
The camera [camera movement].
Examples:
* `[Static shot]`
* `[Push in]`
* `[Pull out]`
* `[Pan left]`
* `[Pan right]`
* `[Tilt up]`
* `[Tilt down]`
* `[Truck left]`
* `[Truck right]`
* `[Pedestal up]`
* `[Pedestal down]`
* `[Zoom in]`
* `[Zoom out]`
* `[Tracking shot]`
* `[Shake]`
Keep the camera movement continuous and appropriate to the action.
## Visual Continuity
Maintain the environment established by `<Picture 1>` throughout the shot.
Preserve its spatial geometry, object placement and overall visual identity.
Do not duplicate, offset, mirror or layer a second interpretation of the environment.
Subject 1 retains the identity established by `<Picture 2>`, `<Picture 3>` and `<Picture 4>`.
Subject 2 retains the identity established by `<Picture 5>` and `<Picture 6>`.
Do not mix facial or physical characteristics between subjects.
## Dialogue
Subject 1 (S1) says:
<d>[English] [SUBJECT 1 DIALOGUE]</d>
Subject 2 (S2) says:
<d>[English] [SUBJECT 2 DIALOGUE]</d>
[If characters speak together:]
Subject 1 and Subject 2 (S1,S2) say together:
<d>[English] [SHARED DIALOGUE]</d>
## Audio
**ambient_sound:** [environmental ambience]
**sound_effects:** [specific diegetic sounds caused by actions]
**non_diegetic_music:** N/A