Cinematográficos · NOVO

Driver Portrait

Você adiciona: Um retrato nítido de corpo inteiro com espaço para movimento cinematográfico. Envie no estúdio.

Recorte a imagem inicial para 16:9; este modelo mantém o enquadramento da imagem enviada.

Inspiração viral

Prévia inspiradora de alta qualidade de cinco segundos. Gere uma tomada de 5s ou 10s, ou planeje um filme de várias tomadas de 30–120s.

PROMPT EXATO DA COMUNIDADE

CHARACTER REPLACEMENT TASK — SINGLE CHARACTER

1. TARGET CHARACTER:
Replace ONLY the single main character in the original video with the character from <<<image_1>>>. Do not replace any other person in the video.

2. CHARACTER PLACEMENT:
The replacement character must occupy the exact same position as the original target character in every frame. Preserve the original character's body placement, scale, proportions, pose, orientation, and position within the frame.

3. IDENTITY LOCK:
Use <<<image_1>>> as a strict identity reference throughout the entire video.

Preserve exactly:
- facial identity
- facial structure
- hairstyle
- skin tone
- body proportions
- clothing
- clothing colors
- overall appearance

The character must remain the SAME person throughout the entire video.

NO face morphing.
NO identity drift.
NO facial changes.
NO blending with the original character.
NO changes to hairstyle, clothing, or body proportions.

4. MOTION PRESERVATION:
Preserve the original target character's movement frame by frame.

The replacement character must perform the exact same:
- body movements
- arm and hand movements
- leg movements
- head movements
- body rotation
- walking or dancing motion
- facing direction
- timing and rhythm

Do NOT create new choreography or alter the original movement.

5. OTHER PEOPLE:
Every other person visible in the original video must remain COMPLETELY UNCHANGED.

Do not replace, modify, duplicate, remove, or regenerate any surrounding person.

6. CAMERA AND FRAMING:
Keep the original:
- camera angle
- camera movement
- framing
- focal length
- perspective
- depth of field
- zoom
- shot duration

Do not crop, reframe, zoom, or stabilize the footage.

7. ENVIRONMENT:
Keep the original environment unchanged.

Preserve the:
- background
- buildings
- floor
- walls
- furniture
- props
- scenery
- objects
- lighting environment

ONLY the designated target character is modified.

8. LIGHTING AND REALISM:
Match the replacement character to the original scene's:
- lighting direction
- light intensity
- color temperature
- highlights
- shadows
- reflections
- ambient lighting
- contact shadows

The replacement must look naturally filmed in the original scene, NOT pasted or composited onto it.

9. SCALE AND DEPTH:
Maintain the exact original size, perspective, distance, and depth of the target character.

Preserve the original occlusion order between the target character, other people, and objects.

10. TEMPORAL CONSISTENCY:
Maintain perfect frame-to-frame consistency.

The replacement character's:
- face
- hair
- skin tone
- clothing
- colors
- body proportions
- identity

must remain stable throughout the entire video.

NO flickering.
NO identity changes.
NO clothing changes.
NO body deformation.
NO texture crawling.

11. FULL COLOR OUTPUT — IMPORTANT:
The final output MUST be a FULL-COLOR VIDEO.

Do NOT convert the footage to:
- black and white
- grayscale
- monochrome
- sepia
- desaturated
- vintage monochrome
- muted color

Preserve natural, realistic colors throughout the entire video.

Maintain the original scene's color information while keeping the replacement character's natural skin tone, clothing colors, and environmental colors accurate.

The final video must look like a normal modern COLOR VIDEO.

12. NO TEXT / NO WATERMARK — ABSOLUTE:
The final video MUST contain absolutely NO text or graphic overlays.

Do NOT add:
- subtitles
- captions
- titles
- logos
- usernames
- labels
- timestamps
- watermarks
- signatures
- AI-generated text
- interface elements
- borders containing text
- any other graphic text

If the source video contains any text or watermark, do NOT reproduce, enhance, or add it to the output. Keep the visual area clean.

13. IMAGE QUALITY:
Preserve the original video's visual quality, sharpness, motion blur, grain/noise, compression characteristics, and realistic visual appearance.

Do not apply:
- cartoon effects
- anime effects
- beauty filters
- artificial skin
- excessive sharpening
- cinematic color grading
- stylization
- visual effects

AVOID:
- replacing other people
- adding people
- removing people
- changing the background
- changing props
- changing camera movement
- changing choreography
- face morphing
- identity drift
- distorted face
- warped body
- extra fingers
- missing limbs
- duplicated limbs
- unnatural joints
- floating feet
- incorrect shadows
- incorrect occlusion
- flickering
- clothing changes
- hairstyle changes
- color shifts
- grayscale
- black and white
- sepia
- desaturation
- cartoon look
- text
- subtitles
- captions
- logos
- watermark
- username
- timestamp
- graphic overlay
- slow motion
- speed ramp
- frame interpolation artifacts
- camera re-framing
- zoom
- crop
- stabilization

OUTPUT:
Generate a FULL-COLOR video with ONLY the designated target character replaced by <<<image_1>>>.

Everything else must remain identical to the source video.

Same resolution.
Same aspect ratio.
Same frame rate.
Same duration.
Same camera movement.
Same choreography.
Same environment.
Same timing.
FULL COLOR.
NO TEXT.
NO WATERMARK.
NO GRAPHIC OVERLAY.

PRIORITY:
1. Preserve the original video.
2. Replace ONLY the designated target character.
3. Preserve the exact identity of <<<image_1>>>.
4. Preserve the original movement and timing.
5. Preserve natural full-color appearance.
6. Do not add any text, watermark, logo, or graphic overlay.
7. Do not modify anything else.

Este prompt exato exige mídia de referência obrigatória ou áudio preservado que este modelo não pode receber. Copie-o para usar em outro lugar.

PROMPT DE GERAÇÃO

Adaptado ao modelo abaixo para criar conteúdo semelhante com sua imagem. Os resultados variam a cada imagem e geração.

Envie uma imagem inicial nítida que siga as orientações sobre o sujeito.
Use este modelo para carregar seu prompt editável de dez segundos e suas configurações.
Gere uma tomada de 5s ou 10s, ou continue no Film Studio para um filme mais longo com várias tomadas.

Custo: 64 cr
Clipe direto

Precisa de um vídeo mais longo?

Planeje o prompt editado como tomadas conectadas de 10 segundos e monte um filme de 30–120 segundos.

Quer ajuda para ajustar este modelo?

Leve o prompt editado para o LLM Chat, onde Prompt writer, Storyboard, Model picker e todas as ferramentas do Studio estão disponíveis.

Abrir no LLM Chat →

Nada é cobrado até você pressionar Gerar. Gerações com falha são reembolsadas integralmente.