Usa el generador de video Veo 3 de CanvasSeed para explorar la planificación de escenas con Veo. Separa el brief visual de tu idea de sonido. Usa los controles de audio solo cuando el modelo seleccionado los exponga, e inspecciona el clip resultante y el sonido. Revisa el modelo disponible, la configuración compatible y la cotización de créditos actual antes de enviar una tarea.
Comienza con la planificación de escenas de video con Veo y un objetivo visual concreto. Describe la acción visual y el encuadre de cámara, añadiendo indicaciones de audio solo cuando el modo seleccionado las admita.
Para una idea basada en referencias, sube material que tengas permiso para usar y comprueba los requisitos del modelo seleccionado. Indica qué debe seguir siendo reconocible y qué debe cambiar.
Planifica el encuadre según el uso previsto: una portada, un concepto de producto o una publicación social. Decide dónde debe mirar el espectador al inicio y cómo termina la acción.
Explora los ejemplos publicados y la biblioteca de prompts de este sitio para entender cómo se relaciona un brief con un resultado. Lee la etiqueta real del modelo, elige una idea compositiva útil y reescribe el prompt para tu sujeto.
Usa la galería de videos publicados para estudiar un resultado real junto con su etiqueta de modelo y el prompt disponible. Observa cómo se desarrollan la acción y la dirección de cámara a lo largo de todo el clip. Las seis ideas de prompt en esta página son briefs sugeridos que puedes adaptar, y no afirmaciones de que esas escenas exactas fueron generadas.
Seedance 2.0
0:00 - 0:02 – A sleek, matte black BMW luxury sedan is parked in a dark, atmospheric studio. Tendrils of white smoke curl around the base of the vehicle and its large multi-spoke wheels, giving it a dramatic, mysterious look. 0:02 - 0:03 – The camera transitions to a rear three-quarter angle of the car as smoke gently drifts around it. The text on the screen reads: "Some things don't ask for attention." 0:04 - 0:06 – A close-up shot focuses on the spinning front wheel with a prominent blue and white BMW logo at its center, showcasing the car's intricate, dark-alloy rim design amidst swirling smoke. 0:07 - 0:09 – The focus shifts to the sharp, modern LED taillight of the vehicle as it illuminates into a bright red glow. The text on the screen changes to: "They command it." 0:10 - 0:12 – A close-up reveals the sleek front headlight assembly, which lights up dynamically with bright white daytime running lights. 0:13 - 0:15 – A macro shot focuses on a textured carbon fiber component of the car. The text reads: "Engineered in silence." 0:16 - 0:19 – The camera pans across the detailed pattern of the dark-finished multi-spoke wheel and high-performance tire. 0:20 - 0:29 – The setting shifts outdoors. The matte black BMW is shown driving smoothly along a wet coastal road during twilight, with ocean waves crashing against rocks in the background. The text on the screen reads: "Then it moves." 0:30 - 0:40 – The vehicle is back in the studio, viewed from a front three-quarter angle under a stark spotlight. A lock icon appears above the text: "Book a private viewing." and a final luxury emblem logo finishes the scene.
Seedance 2.0
Create a 15-second vertical 16:9 UGC-style shampoo advertisement that feels authentic, premium, and social-media ready. A confident young woman stands in a bright, modern bathroom with soft natural morning light. She holds the shampoo bottle close to the camera, smiling warmly as if recommending it to a friend. Her hair is long, healthy, shiny, and naturally bouncy. She speaks directly to the camera: "I've finally found a shampoo that leaves my hair feeling soft, strong, and incredibly shiny. It cleans without drying, smells amazing, and honestly makes every wash feel like a salon treatment. If you're looking for healthier-looking hair, you have to try this!" Storyboard: Scene 1: Close-up of the shampoo bottle in her hand as she introduces it. Scene 2: She dispenses the shampoo into her palm, creating a rich, creamy lather. Scene 3: Beauty shots of her gently washing and massaging her hair, with realistic foam and flowing water. Scene 4: After drying, she flips her silky, glossy hair while smiling confidently at the camera. Scene 5: Final hero shot holding the shampoo beside her face with a clean, premium bathroom background. She smiles and says, "Highly recommend!" The video should have realistic smartphone camera quality, natural handheld movement, genuine facial expressions, smooth transitions, accurate lip sync, premium product visibility throughout, and a polished yet authentic UGC aesthetic. The shampoo bottle label remains sharp, readable, and consistently visible in every product shot.
Seedance 2.0
Cinematic 10-second luxury unboxing video for the "ETERNAL VOW" wedding ring by LUMINA Bridal. Open on a soft, romantic setting with warm ivory silk fabric and delicate rose-gold accents under gentle golden sunrise lighting. A sleek matte cream and rose-gold jewelry box sits centered. The lid opens smoothly with a soft click, revealing a timeless platinum and diamond wedding band nestled in blush velvet. A gloved hand in ivory silk gently lifts the ring. The diamonds sparkle with warm brilliant fire as soft golden particles and elegant light ribbons swirl around it. The ring levitates gracefully, slowly rotating to showcase its perfect round brilliant cut and engraved inner band, while delicate golden lens flares and sparkling highlights dance across the frame. Smooth cinematic camera movement: gentle orbit around the floating ring with a slow upward tilt. Photorealistic details, ultra-premium commercial quality, volumetric lighting, shallow depth of field, luxurious warm rose-gold and ivory color grading, hyper-realistic textures, romantic and timeless mood.
FLUX 3
Seedance 2.0
Main Subject: Young Korean woman, early 20s, natural everyday appearance, faded charcoal-grey sleeveless crop top, loose high-waisted light-wash jeans, black canvas sneakers, black cord necklace, black wavy hair in a messy side ponytail with wispy bangs. Realistic skin texture, minimal makeup, warm and approachable personality. Maintain identical identity, clothing, hairstyle, and appearance throughout. Autumn Leaves Location: Tree-lined residential lane during autumn. 00:00–00:03 She kicks through colorful fallen leaves. 00:03–00:06 Camera briefly loses focus as leaves blow across the lens. 00:06–00:09 She collects a beautiful maple leaf and smiles. 00:09–00:12 Walking beneath orange trees while sunlight flickers through branches. 00:12–00:15 She waves casually before the camcorder stops.
Seedance 2.0
The subject is based on @ image1. Face shape, features, skin tone, damp slicked-back short black hair, black satin cami dress, layered silver necklace, and silver earrings must match @ image1 exactly — no facial distortion allowed. Ignore @ image1's white studio background, frontal standing pose, and bottom text. Even if the image quality is rough, the facial features should remain clearly recognizable. 9:16 vertical format. The entire clip should have the low-quality look of an old iPhone everyday video. Rough noise and compression artifacts persist throughout. Early summer dawn on a rooftop — the sun already harsh, the air still and humid. 0-4 seconds: Hanging wet laundry on a rooftop clothesline, morning light harsh and already bright, wiping sweat from her forehead with her wrist as she reaches up to clip a sheet. 4-8 seconds: Ducking back into the stairwell shade, leaning against the doorframe — a long exhale of relief from the shadow's coolness. 8-12 seconds: Standing selfie angle against the backdrop of damp hanging laundry, low and slightly husky voice saying "…doing laundry before it gets too hot. Yeah." (Korean: "…너무 더워지기 전에 빨래하는 거예요. 네.") 12-15 seconds: Looking out at the hazy city skyline beyond the rooftop railing, a hanging sheet rippling in the wind behind her, the frame trembling slightly, final frame freezes. Camera: 1–1.5 second loose/casual cuts. Old iPhone handheld vertical shooting, slight body sway, autofocus hunting in changing light, harsh sun causing exposure flicker between shade and direct light. Sound: No background music. Ambient city hum as base layer, fabric flapping in wind, distant traffic murmur, faint birdsong, voice close-up but slightly muffled. Style: Uncolored, low-quality phone video texture. No color grading, no film look, no cinematic look, no beauty filter, no CG texture. Preserve authentic skin texture (pores, flyaway hairs, sweat sheen). No merging shots or skipping scenes. No on-screen text.
Seedance 2.0
Ultra-realistic premium product sheet of the Bleu de Chanel perfume bottle, front view, back view, left side, right side, top view, bottom view, 45-degree front angle, 45-degree rear angle, perfectly centered, isolated on a pure white seamless background, soft diffused studio lighting, physically accurate glass material, realistic deep blue tinted bottle, matte black cap, crisp label details, luxury commercial product photography, ultra-sharp focus, photorealistic, 8K, HDR, no shadows touching the frame, no props, no text overlays, no watermark.
Seedance 2.0
A dynamic character concept sheet showcases a "High-Speed Infiltration Unit" in a mid-air grappling pose, utilizing a clean, tech-wear aesthetic. The central figure wears an oversized off-white hooded tech jacket and charcoal tactical joggers, accented by vibrant safety-orange hardware and cabling. Lighting is flat and illustrative with soft ambient occlusion, emphasizing the matte texture of the synthetic fabrics and the hard-surface sheen of the white and orange grappling launchers. The composition is a structured UI layout, featuring a large action splash on the left balanced by expression headshots, dynamic pose thumbnails, and detailed equipment callouts on the right. High-contrast typography and minimalist UI elements frame the subject against a neutral, off-white studio background. * **Style:** Modern anime-style character concept art, tech-wear aesthetic * **Key elements:** Grappling launchers, tactical oversized hoodie, UI layout, expression sheet, technical gear * **Lighting:** Studio softbox lighting, neutral color temperature, subtle cel-shading * **Camera:** Orthographic character sheet, multiple angles, wide layout composition
Seedance 2.0
Style: Ultra-photorealistic 8K IMAX cinematic film. Absolutely no CGI, game-engine appearance, or stylized rendering. Lighting: Natural illumination only. Strong backlit (contre-jour) atmosphere with the camera positioned in shadow. Soft sky glow, window illumination, subtle atmospheric haze, and realistic night ambience. No artificial studio lighting. Color Palette: 60:30:10 ratio (primary / secondary / accent). Camera: Professional cinema camera with vintage anamorphic optics. Physical lens characteristics, authentic depth of field, and natural 180° shutter motion blur. Skin Detail: Hyper-realistic pores, fine facial hairs, natural skin asymmetry, subtle capillary warmth, realistic pore shadows, and physically accurate light interaction. Performance: Feature-film quality acting with authentic micro-expressions, tiny pauses before reactions, believable eye movement, natural breathing, and emotionally grounded performances. Characters remain in motion or reacting naturally. Physics: Real-world gravity and momentum. Every object carries believable weight with accurate contact shadows and inertia. Composition: Rule of thirds combined with golden-ratio framing. Every shot contains purposeful movement from the first frame. Continuity: Wardrobe, environment, props, and character appearance remain perfectly consistent throughout every cut. Technical: Native 24 fps cinematic motion, ultra-high-resolution detail, completely stable camera movement. Audio: Ambient environmental sounds only. No music. No narration. No subtitles. Characters LEAD WOMAN: Wearing a tailored pastel-blue pantsuit over a blush satin camisole, delicate gold necklace, pointed rose-colored heels, long honey-blonde hair worn loose. She radiates excitement, confidence, and genuine happiness, walking with effortless energy. WINDOW COUPLE: A man and woman in their late twenties wearing elegant retro-inspired evening clothing, seated across from one another inside a classic chrome diner. Their posture is reserved and distant at first. Setting Nighttime in a dense downtown district filled with vintage architecture, towering buildings, glowing storefronts waiting in darkness, and reflective rain-washed pavement. The central rule of this world: Every source of light remains completely dark until the woman reaches it. The instant she draws alongside a light source, it awakens naturally and remains illuminated behind her. The path ahead is always dark, while everything behind her becomes increasingly alive, creating the feeling that she is bringing the city to life with every step. CUT 1 Fast side-tracking shot on a 35 mm lens, moving alongside her before craning upward toward a tall vertical theater sign. As she walks beneath an apartment building, each individual window illuminates precisely as she passes below it, floor after floor, perfectly synchronized with her footsteps. A towering vertical marquee, previously unlit, begins activating from the bottom upward as she reaches it. Each glowing letter appears in sequence, building into an original fictional title in vibrant magenta neon. She instinctively looks upward, smiles, and lets out a quiet laugh as the final letter ignites. CUT 2 Medium shot through the large diner window using a 50 mm lens. The camera slowly pushes inward through reflections on the glass. An elegant cursive neon sign above the window remains dark until she reaches it. The moment she passes, the sign hums softly to life, casting a cool turquoise glow across the chrome interior. Focus shifts smoothly through the glass toward the seated couple. As the light reaches them, they simultaneously look up, make eye contact for the first time, and exchange a small, genuine smile, subtly changing the atmosphere between them. CUT 3 Wide 24 mm crane shot slowly rising while drifting in a gentle arc around her. She crosses a deserted city intersection. Every nearby source of light responds instantly to her movement. Traffic signals awaken. Storefront signs illuminate. Vintage arcade lights flicker on. Hotel arrows begin glowing. Street lamps awaken one after another. The activation spreads outward like ripples across every street connected to the intersection. Behind her, the city becomes vibrant and alive. Ahead of her, complete darkness remains until she arrives. Halfway across the intersection she turns briefly, taking in the glowing world she has created. CUT 4 An aerial camera accelerates upward from directly overhead into a sweeping skyline view. Below lies an almost entirely dark city grid. Only one brilliant ribbon of illuminated streets cuts through the darkness. At the end of that glowing path, the tiny figure continues walking. With every step, new lights ignite exactly where she is. The illuminated trail grows continuously behind her while the city ahead remains untouched. As the camera reaches the skyline, her path arrives at the base of a skyscraper. A massive rooftop sign that has remained dark suddenly flickers to life, gradually forming an elegant original cursive wordmark in luminous rose-colored neon. It becomes the brightest light visible anywhere in the city as the film ends.
Seedance 2.0
请围绕“Codex 5.6 的视频制作能力”创作并渲染一条中文横屏短片。 这条视频不要只罗列工具名称，而要让观众直观看到 Codex 如何借助4个 Skill 完成不同的视频任务： HyperFrames 负责创意策划、动态设计、画面合成和最终渲染； Remotion 负责通过 React 精确控制画面并批量生成视频； video-use 负责整理、识别和剪辑已有的真人素材； Seedance Prompt Skill 负责把创意转换成包含镜头运动、人物动作、对白和声音设计的电影感提示词。 视觉采用明亮的白紫配色，画面干净、清晰、有广告片质感。可以使用斜向移动的电影画面、打字动画、流程节点和动态界面，让整条视频始终保持运动感。 4个 Skill 使用统一但不完全重复的视觉结构：左侧说明名称和核心能力，右侧展示对应结果。HyperFrames 和Remotion 使用真实案例；video-use 与 Seedance 使用重新设计的动态流程演示。 我提供的历史视频只用于帮助你理解我的审美偏好，不要复刻其中的文案、镜头、排版或动画。请重新设计画面表 达，保证最终作品具有独立的视觉创意。 完成后检查文字是否清楚、图片是否锐利、布局是否拥挤、画面是否重复、动画是否流畅，以及模拟内容是否被正确标注。确认无误后输出高清 MP4。
Seedance 2.0
A cinematic, high-intensity fight sequence in a dark, gritty industrial warehouse at night with dim overhead lights, green peeling walls, chain-link fences, metal scaffolding, and scattered debris. A beautiful East Asian woman with long flowing black hair, visible facial scratches and intense determined expressions, is engaged in a raw, dynamic hand-to-hand combat against a male attacker in dark clothing. She wears a delicate, flowing baby pink dress (instead of her original jacket) that contrasts dramatically with the violent action — the soft pink fabric billows, swirls, and tears slightly during the fierce movements, hair whipping wildly. The sequence includes: close-ups of her fierce face and flowing hair, powerful punches and kicks with motion blur, grappling, throwing the man down, dramatic spins, hair flips, kneeling over the defeated attacker, then rising victoriously while breathing heavily. Moody teal-green color grading, realistic particle effects like dust and fabric motion, intense choreography, dramatic camera angles (handheld, tracking shots, low angles), high detail, filmic lighting, 4K cinematic quality, emotional and empowering tone.
Seedance 2.0
BROADCAST TYPE: Live business news report covering a sudden stock market crash during trading hours. ON-AIR REPORTER: Female journalist, early 30s. Sharp blazer, earpiece, handheld mic with financial network logo. Composed but visibly tense. LIVE BROADCAST TIME: 11:20 AM. Bright daylight, glass-walled trading floor visible behind glass partition. REPORT LOCATION: Outside a stock exchange building. Digital ticker boards flashing red numbers. Traders visible through windows, gesturing anxiously. Passersby stopping to check phones. BROADCAST CAMERA: Steady tripod-mounted main shot with a secondary handheld cutaway camera. Occasional quick zoom to ticker board numbers dropping. Reflections of red digital numbers on glass. LIVE SEGMENTS: Opening: Reporter explains the sudden index drop and market panic. Update: Camera cuts to trading floor screens showing plunging graphs. Eyewitness Moment: A trader stepping outside briefly comments on the chaos. Situation Change: Ticker shows a fresh plunge, reporter reacts live to the number. Closing: Reporter signs off, noting market will be watched closely into afternoon. LIVE AUDIO: Muffled trading floor shouting, city traffic, notification pings from nearby phones, reporter's controlled voice, no music. BROADCAST REALISM: Genuine tension, traders ignoring camera, real-time number changes driving reactions, natural interruptions from passersby glancing at screens.
Idea de prompt: Una zapatilla sin marca descansa en un estudio minimalista mientras una pequeña gota golpea la suela y la cámara revela el perfil lateral. Mantén el producto legible; inspecciona el impacto y la forma a lo largo de la toma.
Idea de prompt: Un sendero costero se curva hacia el mar al amanecer, con un movimiento lento de cámara hacia adelante y un suave movimiento de la hierba cerca del borde del encuadre. Juzga la profundidad y el ritmo de la escena antes de pedir más actividad ambiental.
Idea de prompt: Una cuchara se mueve a través de un tazón de sopa mientras el vapor se eleva suavemente. Usa un primer plano fijo, luz cálida y un fondo simple. Inspecciona el contacto, el comportamiento del líquido y los cambios no deseados en el tazón.
Idea de prompt: Usa una imagen compatible de un salón y describe un sutil avance de cámara hacia la ventana, manteniendo estables los muebles y la geometría de la habitación. Compara la referencia con varios fotogramas en lugar de revisar solo la apertura.
Idea de prompt: Describe un tren llegando a un andén tranquilo. Si se ofrece audio, solicita el ruido suave de las ruedas y el sonido ambiental de la estación por separado. Escucha el resultado real; no asumas que cada modelo o modo produce sonido.
Idea de prompt: Una florista coloca un tallo en un jarrón, grabado verticalmente con el arreglo centrado y suficiente espacio arriba para el texto posterior. Revisa la acción completa y el recorte previsto antes de añadir un titular en tu editor.
Comienza con la planificación de escenas de video Veo. Nombra el sujeto, el entorno y el propósito, luego describe la acción visual y el encuadre de cámara, añadiendo indicaciones de audio solo cuando el modo seleccionado las admita.
Selecciona un modelo de video disponible y lee los controles que se ofrecen para la tarea. Usa una referencia solo cuando esa entrada sea compatible. Elige el encuadre para el lugar donde aparecerá el resultado y deja espacio para cualquier recorte o texto que agregues después.
Comprueba el presupuesto de créditos actual después de elegir el modelo y la configuración de salida. Guarda tu brief y los controles seleccionados para poder comparar intentos posteriores. Envía la tarea cuando las entradas coincidan con tu plan y espera su estado en lugar de enviar repetidamente la misma solicitud.
Abre el resultado completado en History o Assets e inspecciona el encuadre, el movimiento y cualquier audio realmente devuelto del clip. Mira el clip completo y compara su inicio, su parte media y su final.
Usa la etiqueta de modelo y los controles reales que se muestran en el generador. El enfoque aquí es la planificación de escenas de video Veo; la compatibilidad de entrada, las opciones de salida y los costos de créditos aún dependen de tu selección. Compara modelos usando el mismo brief cuando sus entradas lo permitan.
Comienza con los créditos incluidos, compara planes y agrega más créditos cuando los necesites.
Cargando precios…
El pago seguro está disponible. Para consultas o asistencia, contáctanos en support@canvasseed.com.
Sí. No se requiere registro. La generación usa créditos; revisa el presupuesto y tu saldo antes de enviar. Los créditos de invitado y el historial están vinculados a este navegador, así que mantén las cookies activadas. Puedes comprar paquetes de créditos únicos elegibles cuando el pago esté disponible. Las suscripciones requieren iniciar sesión.
Usa CanvasSeed para explorar la planificación de escenas Veo. Selecciona un modelo que admita tu tarea prevista y revisa sus opciones de entrada actuales antes de generar. Comienza con la planificación de escenas de video Veo.
Comienza con un sujeto, un entorno y una dirección visual específicos. Idea de prompt: Una zapatilla sin marca descansa en un estudio minimalista mientras una pequeña gota impacta la suela y la cámara revela el perfil lateral. Mantén el producto legible; inspecciona el impacto y la forma a lo largo de la toma. Elige el encuadre y la duración entre las opciones que realmente se muestran para el modelo. Trata esto como un brief inicial: adapta los detalles a tu propio proyecto, luego cambia una instrucción después de inspeccionar el resultado.
Elige un flujo de trabajo que admita entradas de imagen. Describe la acción y el movimiento de cámara que deseas, y luego revisa el resultado en busca de cambios en el sujeto original.
El generador muestra un presupuesto para el modelo y la configuración seleccionados. Revisa ese presupuesto antes de enviar; cambiar la selección puede cambiar el coste.
La etiqueta real del modelo de cada ejemplo identifica el modelo utilizado. Los ejemplos pueden usar modelos distintos del que selecciones para una tarea nueva.
Usa los modelos que se enumeran actualmente para la tarea o contacta con soporte sobre la disponibilidad. Un encabezado de página o un ejemplo publicado no garantiza que un modelo pueda seleccionarse ahora.
Comienza con uno de los briefs sugeridos para la planificación de escenas de video Veo, adáptalo a tu sujeto y elige una tarea compatible. Revisa la cotización antes de generar, luego inspecciona el resultado contra tu plan original.