There is a moment in Fidā when the image of a single tear running through orange paint becomes devastating.
It opens with striking, symbolic beauty. A close-up of a woman with shaved head and intricate orange markings on her dark skin. Then the film moves between two worlds: a sacred, torch-lit underground chamber where hooded figures stand in ritual, and a man kneeling on red earth, hands clasped, waiting. The cinematography is rich and cinematic, full of dust, dramatic light beams, and heavy atmosphere. Everything feels ancient, solemn, and intentional.
As the film progresses, the images layer with quiet intensity. The woman's tear falls. The man sits in quiet acceptance. A hooded figure approaches with a blade. A circle of watchers stands in silence. There is almost no dialogue. The story is told entirely through composition, performance, and atmosphere. Yet you understand exactly what is happening: a willing sacrifice, an act of devotion, a life given for something larger.
What makes Fidā special is how confidently it embraces restraint. Studio.13 trusts the visuals to carry the emotional weight. The orange markings, the red earth, the contrast between stillness and ritual tension, every choice feels deliberate. The film never rushes. It lets silence and image do the work that most shorts try to force with voice-over or exposition.
In just 84 seconds, it creates a complete ritual world that feels lived-in and spiritually heavy. The final title card arrives like the closing of a prayer.
Most AI films right now are busy showing what they can generate. Fidā is interested in something harder. It uses the tools in service of tone, symbolism, and emotional precision. It proves a short film can feel sacred, ancient, and deeply human even when made with modern technology.
That kind of visual discipline and emotional clarity is rare.
That is what Wondra looks for.
Film Details
Runtime1:25
Source
Hosted on Wondra
Modelblender, comfyUI, Davinci resolve
Tools
sci-fi-dystopianretro-vibesurreal-dream
Generation Prompt
This is 3% prompting and 97% pure pixel control from step diffusions by using 3d blocking and finishing in davinci resolve studio in an ACEScct color space, then Foley, the. Timeline edit. This is the result of using control nets like open pose, canny edges, Depthmaps for relight, height maps + depth maps for a 3d style diffusion in a 2d space where there's maximum control of camera position and frame composition, then inference in another custom workflow in comfy using Kline03 API in the pipeline. All work done in EKR 32 bits of pixel data.