This dynamic prompt is designed to generate a highly cinematic, 8K-resolution adventure scene featuring a muscular man standing waist-deep in water at night, illuminated by a flaming torch. Because the prompt explicitly demands 100% facial fidelity to an uploaded image (“use a referência da imagem carregada para o rosto, seja 100% fiel”), you must use an AI workflow that supports character referencing.
Here is a step-by-step guide on how to execute this prompt effectively.
Step 1: Select a Capable AI Generator To achieve accurate facial representation, you need a tool with advanced image referencing. Midjourney (using the --cref parameter), Stable Diffusion (utilizing tools like IP-Adapter or FaceID), and Leonardo AI (using Image Guidance) are the best platforms for this specific task.
Um homem musculoso, atlético e com musculos definidos e volumosos, sem camisa (use a referência da imagem carregada para o rosto, seja 100% fiel) com cabelo escuro e ondulado e abdominais definidos, vestindo shorts de cor clara, em pé com água na altura da cintura em água escura à noite. Ele está segurando uma tocha flamejante em sua mão direita, iluminando seu corpo com um brilho quente e tremeluzente. Ele está olhando para o lado com uma expressão determinada. O fundo é um céu noturno escuro e sem lua e água escura. Iluminação dramática, cinematográfico, hiper-realista, brilho de tocha, tema de aventura, 8k
Step 2: Upload the Perfect Reference Image Choose a high-quality, clear portrait of the subject. Because the prompt dictates that the character is “looking to the side with a determined expression” (olhando para o lado com uma expressão determinada), uploading a reference photo where the subject is looking slightly off-camera will yield the most natural and seamless blend.
Step 3: Input the Prompt The provided prompt is written in Portuguese. Modern AI generators like Midjourney and DALL-E 3 process Portuguese seamlessly, so you can paste it exactly as written. However, if you are using an older or English-specific model (like some Stable Diffusion checkpoints), translate the text to English first to ensure maximum accuracy:
“A muscular, athletic man, shirtless, dark wavy hair, defined abs, wearing light shorts, standing waist-deep in dark water at night. Holding a flaming torch in his right hand, warm flickering glow on his body, looking to the side, determined expression. Dark moonless sky, dramatic lighting, cinematic, hyper-realistic, adventure theme, 8k.”
Step 4: Generate and Fine-Tune Generate the image. The visual impact of this prompt relies heavily on the contrast between the dark environment and the glowing torch. If the resulting image is too bright, simply add negative prompts like “daylight, bright sky, overall lighting” to force the AI to respect the dramatic, shadowed atmosphere.