ThiArt AI @thitechart

SHOT 1 (0-6s): low tracking shot, camera mounted just above road level, chasing a vehicle from behind and slightly to the left. Slight handheld instability, heavy motion blur, anamorphic lens, shallow depth of field, faint lens flares. SUBJECT: a matte black angular hypercar, exposed rear wheels, tall rear wing, twin amber taillights, blue ion thrusters glowing at the rear diffuser. It speeds down a wet elevated highway, throwing a wide spray of water from its tires. SETTING: dense cyberpunk megacity at night in torrential rain. Towering high-rises on both sides covered in holographic billboards and neon kanji signage. Heavy volumetric fog, visible rain streaks, standing water on black asphalt acting like a mirror. Sparse traffic ahead. A small flying transport craft crosses low above the highway from right to left. LIGHT: no sun, city-lit only. Dominant desaturated teal-cyan ambient wash from the billboards, punctuated by warm orange and amber point sources — taillights, sodium street lamps, distant window glow. Crushed blacks, low overall exposure, high atmospheric haze so distant buildings fade into grey-green mist. Reflected light smears along the wet road surface. SHOT 2 (6-10s): the camera pulls back and rises into a wide centered view. The highway leads into a colossal dark structure — a gate cut into the base of a megastructure — with a towering vertical shaft of glowing cyan-teal light and billowing mist pouring from the opening. Two cars ahead drive toward it, silhouetted against the glow. Symmetrical composition, guard rails converging to the vanishing point. GRADE: cinematic teal-and-orange, film grain, slight chromatic aberration, 2.39:1 letterbox feel. AUDIO: torrential rain, tire spray on wet asphalt, a deep electric turbine whine from the car rising in pitch, distant city hum, low sub-bass drone as they approach the gate. CONSTRAINTS: continuous camera move, no hard cuts, no dialogue, no subtitles, no on-screen text.

Cyberpunk car scene Prompt in the comments

@mnowakdesign Nice! is it AI video generation or After Effects?

@SimplyAnnisa @TapNow_AI Hmm... I'm getting hungry just looking at it 😋

@yoland_yan wow are you using a free model?

Nice! I tried this variations with @Hailuo_AI H3 SHOT: 1990s live news broadcast, 4:3 interlaced SD video, heavy chroma bleed, tape noise, shoulder-mounted camera, handheld and unstable. Harsh camera light against a flat grey sky. Lower-third chyron: "LIVE — WKBR 9 — COASTAL WARNING". SUBJECT: a woman field reporter in a soaked yellow rain jacket holding a station-branded microphone, standing knee-deep in fast brown water on a flooded seaside street. ACTION: she holds full professional composure, steady broadcast cadence, straight to camera: "—authorities are asking residents to move inland, we're being told that—". Behind her people run past and a parked car slides sideways in the current. Her eyes flick to something off camera. She stops mid-word. A wall of water fills the frame. The camera is knocked down and goes under, tumbling into churning brown water and darkness. AUDIO: wind on the mic, sirens, distant shouting, her voice steady then cut off, a low roar building, then everything muffled underwater. CONSTRAINTS: one continuous take, no cuts, no subtitles, no on-screen text other than the chyron.

SHOT: 1990s live news broadcast, 4:3 interlaced SD video, heavy chroma bleed, tape noise, shoulder-mounted camera, handheld and unstable. Harsh camera light against a flat grey sky. Lower-third chyron: "LIVE — WKBR 9 — COASTAL WARNING". SUBJECT: a woman field reporter in a soaked yellow rain jacket holding a station-branded microphone, standing knee-deep in fast brown water on a flooded seaside street. ACTION: she holds full professional composure, steady broadcast cadence, straight to camera: "—authorities are asking residents to move inland, we're being told that—". Behind her people run past and a parked car slides sideways in the current. Her eyes flick to something off camera. She stops mid-word. A wall of water fills the frame. The camera is knocked down and goes under, tumbling into churning brown water and darkness. AUDIO: wind on the mic, sirens, distant shouting, her voice steady then cut off, a low roar building, then everything muffled underwater. CONSTRAINTS: one continuous take, no cuts, no subtitles, no on-screen text other than the chyron.

A perfect disaster simulation 😱 @MiniMax_AI H3 Prompt in the comments 👇

@jerrod_lew for image generation, do you prefer GPT Image 2 or Nano Banana?

Which model's result did you like the most?

@JesusPlazaX really nice! I'd like to know the shortfilm

Fisheye doorbell camera footage, night vision infrared, grainy, timestamp overlay "03:14 AM" in the top right corner. A large brown bear steps onto a suburban front porch and picks up a cardboard delivery box. It turns the box over, squinting at the shipping label, and mutters in a low gravelly voice: "...this isn't even my address." The bear carefully sets the box back down exactly where it was, pats it once, and walks off the porch into the dark. Static camera, no camera movement. Ambient audio: crickets, distant traffic, cardboard scraping. No subtitles, no on-screen text other than the timestamp.

Testing night vision footage from a prompt @FlowbyGoogle veo3.1 fast run it yourself. prompt in the comments ↓

@DaNel0325 Nice! This is my try 😀

I added some style to this shoe of the future @FlowbyGoogle Omni

𝔻𝕒𝕟𝕖𝕝 @DaNel0325

a month ago

Shoes from the future🔥

3K 3K 29K 10.7M 6K

@MrDasOnX great job!

Finally, Hulk! 😂

@EHuanglu I tried this 😀

Style: Ultra-photorealistic, cinematic, emotional drama, extremely expressive facial animation, Hollywood-quality acting, IMAX, shallow depth of field, 85mm lens, natural lighting, subtle handheld camera, realistic fur simulation, physically accurate eye reflections, no cartoon look, no exaggerated animation. Scene Description: The scene begins with the first image as the exact starting frame. The black fluffy dog is facing another identical dog at extremely close distance. His right paw is tightly clenched into a fist, raised and ready to throw a punch. His expression is a complex mix of anger, frustration, sadness, and hesitation. His eyes are locked onto the other dog with intense emotional conflict. The camera remains in a dramatic close-up throughout the sequence, prioritizing subtle facial expressions over body movement. Timeline 0.0 – 1.5 seconds The dog remains perfectly still with his fist raised. His breathing is heavy and visible through subtle movement in his chest and muzzle. His eyebrows remain tense, his jaw tight. His eyes burn with anger while also showing emotional pain. 1.5 – 3.0 seconds Tiny facial micro-expressions begin to appear. His eyes slightly soften. His ears lower almost imperceptibly. His fist remains clenched but begins to tremble slightly. His breathing slows. 3.0 – 5.0 seconds The internal conflict becomes obvious. His eyes become glossy with emotion as if holding back tears. His gaze briefly shifts downward before returning to the other dog. His fist slowly loses tension although it is still raised. 5.0 – 6.5 seconds He exhales deeply. The anger fades from his face and transforms into sadness and compassion. His paw slowly opens from a fist into a relaxed paw. His shoulders visibly relax. 6.5 – 8.0 seconds Instead of striking, he slowly lowers his paw while maintaining eye contact. His expression changes completely from aggression to vulnerability. The tension dissolves. 8.0 – 10.0 seconds He gently steps forward and wraps both front paws around the other dog in a heartfelt embrace, perfectly matching the supplied final image. The hug feels genuine, emotional and relieving. His eyes close naturally as he hugs. His breathing becomes calm. The second dog immediately accepts the embrace without resistance. The camera performs a subtle cinematic push-in during the hug, ending with the exact composition of the provided final image. Hold the embrace for the final second, emphasizing warmth, forgiveness and emotional release. Negative Prompt No violence. No actual punch. No sudden movements. No cartoon animation. No comedy. No smiling before the hug. No exaggerated facial distortion. No morphing artifacts. No extra limbs. No changes in fur color. No camera cuts. No camera shake. No text. No subtitles. No unrealistic physics. Preserve both dogs' identity and facial features throughout the sequence. The emotional transition must feel slow, believable and deeply cinematic, driven primarily by subtle eye movement, breathing, and facial micro-expressions rather than large body motions.

Trying to do something similar 😁

el.cine @EHuanglu

a month ago

..me trying to make them feel what I felt

19 39 438 41K 204