Your Next Opponent Is Yourself

Model N/ACreativeOutfit ChangeVFXBlack & White10s
The full recipe behind the viral Douyin hit 'Your Next Opponent Is Yourself': two contrasting photos as reference, a deep-black cool-grey dream palette with Tyndall volumetric light, all hard cuts. The character-consistency constraints in this prompt are worth copying.
PROMPT · Prompt
Create with AI
The viral Douyin video tutorial 'Your Next Opponent Is Yourself' AI-generated video + minor post-processing in Jianying (for subtitles) Using Tibo as an example, first find two photos of the PK characters, then input the prompt: Surrealist cinematic concept short film, 10fps, slight motion blur to enhance the dreamlike feel, overall deep black cool-grey tone, high contrast interplay of light and dark, large areas of dark negative space, empty, lonely, quiet yet oppressive, clear Tyndall volumetric light throughout, fine floating dust in the air, cinematic film grain texture, authentic cinematographic quality, all shots are hard cuts, no fade in/out or soft transitions, crisp and decisive pacing. Character consistency must be strictly maintained: 【@Image1】the boy's facial features, face shape, hairstyle, hair color, sense of age, and overall appearance must strictly match uploaded Image 1, without altering his identity characteristics, maintaining natural boyish proportions in full-body shots; 【@Image2】the adult man's facial features, face shape, eye shape, nose shape, lips, hairstyle, and identity must strictly match uploaded Image 2, especially close-up facial shots which must be highly consistent, not beautified into a different face, bare-chested, with clearly strong yet natural pectoral, abdominal, shoulder and back muscle lines, wearing off-white draping wide-leg trousers, barefoot, overall in a classical statue-like still pose, with a dignified, cold, restrained gaze carrying a sense of looking down from above. Total duration 10 seconds, four shots in continuous narrative. Shot 1, 0-5 seconds, medium shot, fixed camera with extremely slight handheld shake, boundless pure dark void, only a vertical spotlight beam from above. 【@Image1】the boy slowly walks toward the camera from deep in the frame, his full body gradually entering the light, glancing slightly left and right while walking, expression lost, alert, searching, the top light is warm but sharp-edged, forming clear rim highlights on his hair, shoulders, and body edges. As the boy reaches the center of the frame, the top light beam suddenly and sharply intensifies, causing a brief cinematic overexposure; the boy instinctively raises his forearm to shield his eyes, head slightly turning away, body slightly hunching, fabric producing realistic pulling and creasing with the motion. After the intense light peaks, it gradually fades, the boy slowly lowers his arm, squinting to adjust to the light, and the moment his arm fully drops and his vision clears again, hard cut. No dialogue, ambient sound is the light friction of bare feet on the ground and fabric rustling, a brief low electric hum occurs at the moment the intense light bursts, then quickly fades to silence. Shot 2, 5-7 seconds, slightly upward-angle wide-angle long shot, camera positioned between the two figures, moving extremely slowly forward, boundless pure dark void, the ground is a gently undulating matte grey dune occupying only about one-fifth of the lower frame, large dark negative space surrounds the figures. 【@Image1】the boy is on the left side of the frame, closer to camera, smaller in stature, stops walking, slowly raises his head to look toward the far right of the frame; 【@Image2】the adult man is on the right side of the frame, far in the distance, standing still facing forward, bare-chested, off-white draping wide-leg trousers, barefoot, the slightly upward wide-angle lens slightly stretches his body proportions, making him appear unusually tall and imposing, like a statue suddenly appearing in the darkness. A large expanse of pure black void is kept between the two, creating strong distance and confrontational tension. A hard slanted light beam pierces the darkness from the upper right, forming an extremely clear Tyndall effect, with fine dust slowly swirling within the light column. 【@Image2】the adult man is outlined by cool-toned side backlight, his pectorals, abdominals, shoulders and collarbone forming sharp sculptural light-and-shadow structure; 【@Image1】the boy's body edge is outlined by a thin silver-white rim light. Their elongated shadows both extend toward the bottom left of the frame; the moment the boy looks up and clearly sees the adult man, hard cut to the adult man's face. Ambient sound is a low void ambient hum, with extremely faint sound of sand sliding. Shot 3, 7-8.5 seconds, close-up of 【@Image2】the adult man's face, framed from the shoulders up, 85mm cinematic portrait lens feel, shallow depth of field, camera positioned slightly below eye level, creating an extremely slight low angle. Must strictly maintain 【@Image2】's original identity and facial consistency, focusing on locking the face shape, brow bone, eye shape, nose bridge, lips, jawline, hairstyle and hair color, without changing his age, without generating a different face. The adult man is almost completely still, hard cool light slants across his face from the upper right, illuminating only about half of his face, while the other half gradually sinks into the pure black background; the brow bone, nose bridge, cheekbones and jaw edge form sharp highlights, skin retains realistic texture and fine pores, no skin smoothing. His eyes are initially in slight shadow, then extremely slowly lift his gaze, looking past below the camera, as if looking down and scrutinizing the boy in front of him. No head turn, no speaking, no exaggerated expression, the emotional change is conveyed entirely through the eyes, the gaze cold, dignified, restrained, without anger yet producing extremely strong pressure. Fine dust in the light column drifts slowly past the foreground of the face, the extremely shallow depth of field turns a small amount of dust into soft blurred light spots, the camera performs an almost imperceptible slow push-in, finally settling near the eyes, and the moment the gaze is fully locked, hard cut. Ambient sound almost entirely disappears, leaving only one extremely low, deep heartbeat pulse. Shot 4, 8.5-10 seconds, side-on eye-level camera position along the line connecting the two figures, medium focal length lens, full-body symmetrical composition, continuing the boundless pure dark void and matte grey dune ground. 【@Image1】the boy is on the left, 【@Image2】the adult man is on the right, both fully in frame, forming an almost perfectly symmetrical visual relationship left and right, the adult man noticeably taller and larger, the boy appearing small. A slanted Tyndall light beam from the upper right pierces through the center of the frame, passing between the two and enveloping both in the same field of light, the light column becoming the sole visual element connecting the two figures. The adult man maintains a completely still statue-like standing posture, sharp side light sculpting the muscular contours, collarbone and body outline of his bare upper body; the boy's entire body is wrapped in a thin silver-white rim light, separating his youthful figure from the darkness. Neither moves nor speaks, the boy looks at the adult man, the adult man's gaze slightly downward, forming a clear contrast in height, age, and presence, their elongated shadows both trailing toward the bottom left of the frame. The frame is almost completely still, only the fine dust in the light column slowly drifting. One final deep low-frequency heartbeat sounds, and at the moment the heartbeat ends, the image and sound stop in sync, cutting directly to black. Throughout the entire film: no subtitles, no voiceover, no dialogue, no slow dissolve transitions, no drifting of facial features, no changing the identities of 【@Image1】and 【@Image2】. Posted 10:11am · August 26, 2026 · 1,442 views 1 6 6
✍️ Editor’s Notes

这条走的是分镜脚本路子,四个镜头卡着秒数写,机位、光效、动作、音效在每一镜头里全部锁死,等于把导演分镜表直接喂给模型,「丁达尔体积光」「硬切」这类词反复出现就是为了摁住转场别自由发挥。双人物一致性靠「@图片1」「@图片2」编号锚点撑着,比堆形容词稳当,但代价是提示词长到近千字,对模型的长文本理解和多主体绑定能力都是硬考验,跑到后半镜头很容易脸部漂移。没写model,大概是留了个坑等强模型来试刀。

Create with AI
Related Works
×