Short answer
The template is a 30-second, 4:3 hangar performance with seven fixed shots and two required seats. The set, trooper rows, camera moves, cuts and 1995 song bed are locked. The Rapper and Singer faces, hair, skin, body types and clothing come from two uploaded photos. The original performers’ faces and robes are discarded.
Make my videoWhat changed
- — Launched the two-photo, 30-second 4:3 Gangsta’s Paradise AI generator.
Gangsta’s Paradise AI template breakdown
The scene is fixed so the casting can be the only variable. Here is exactly what the template keeps, what it replaces, and why both seats are mandatory.
Published 2026-11-09
A template is not a filter. It is a fixed film with two holes in it. The white hangar, the starfield windows, the glossy floor, the rows of white-armored troopers, the one black handheld mic and the seven cuts are the film. The two portraits fill the only two roles that matter.
The set
- Vast white sci-fi hangar with tall columns and a glossy pale floor.
- Tall windows showing deep space and stars.
- Cool white key light with a faint warm rim; no stage haze, no coloured wash.
- Rows of white-armored troopers standing at attention along the back wall.
- One black handheld microphone, one hand, five fingers.
The seven-shot timeline
| Time | Shot | Singer |
|---|---|---|
| 0.0–5.8s | Rapper medium-close, pulls to full body | Not in frame |
| 5.8–11.8s | Singer crosses left to right; push to Rapper face | Walks in and out, silent |
| 11.8–15.0s | Wide, both walk toward camera | Walks, silent |
| 15.0–18.1s | Rapper medium-close, points at camera | Behind, silent |
| 18.1–23.1s | Wide down the hangar, push to raised arm | Distant, silent |
| 23.1–24.2s | Rapper close-up | Behind, silent |
| 24.2–30.0s | Trooper choir, foreground two-shot | Enters ~27s, sings chorus |
What is fixed and what is replaced
| Element | Fixed or replaced |
|---|---|
| Hangar, windows, floor, lighting | Fixed |
| Trooper rows and final choir | Fixed |
| Camera moves and seven cuts | Fixed |
| 4:3 frame, 24 fps, 30 seconds | Fixed |
| 1995 song bed | Fixed |
| Rapper face, hair, skin, body, clothes | Replaced from photo 1 |
| Singer face, hair, skin, body, clothes | Replaced from photo 2 |
| Original performers’ faces and robes | Discarded |
Why both seats are required
The format is a two-hander. The Rapper carries the verse; the Singer’s silent walk and late chorus are the payoff. A one-photo render would either duplicate one face or remove the second role entirely. Neither is the trend. That is why the studio rejects a submission with fewer than two photos and why there is no auto-filled second person.
Why the film is split into two 15-second calls
Wan accepts at most 15 seconds of reference video per call and a reference-plus-output total of 30 seconds. The 30-second reference is therefore cut at 15.0 seconds, the one hard cut that also starts a new musical section. Each half is generated against its own reference and prompt, then stitched frame-exactly and scored. The result is still one continuous 30-second MP4 to the viewer.
Output and price
480p is 640×480 for 7 credits; 720p is 960×720 for 14; 1080p is a 720p master upscaled to 1920×1440 for 22. All three are 30 seconds, 4:3, 24 fps, no watermark, with a silent copy. See pricing for the packs and photo requirements for the two portraits.