AI 3D for Film Pre-Vis (vs Maya / Houdini Storyboards)
Film pre-vis (pre-visualization) is the storyboard-meets-blockout phase before shooting. Directors use it to plan camera moves, blocking, and timing. Maya and Houdini are the studio standards. Yugma fits the early-draft phase.
# The pre-vis pipeline today
- Storyboards (paper sketches, ~1 day per sequence).
- Maya animatic (rough 3D blocking, ~1 week per sequence).
- Houdini effects pre-vis (for VFX-heavy shots, ~2 weeks).
- Final pre-vis (cleaned 3D, ~3-4 weeks per sequence).
Total: 2-6 weeks per sequence pre-vis. The first 1-2 weeks are pure scene-construction overhead.
# With Yugma in front
- Storyboard + initial 3D blocking (Yugma, ~1 day).
- Cleaned animatic (Maya, ~3-4 days).
- Effects pre-vis (Houdini, ~1-2 weeks).
- Final pre-vis (~3 weeks).
Total: 4-5 weeks. Yugma compresses the first week into a day.
# Workflow example: chase scene draft
Director: "Chase scene through a small medieval European town, narrow cobblestone streets, two-story timber houses, late afternoon light, market stalls in the way."
Pre-vis artist types it into Yugma:
"Medieval European street, cobblestone, narrow 5m wide. Two-story timber houses on both sides for 80m. Market stalls every 15m alternating sides. Late afternoon long-shadow lighting. Two protagonist mannequin placeholders at the south end, two pursuer placeholders at north end."
90 seconds, the scene is composed. Director iterates on framing in chat: "add three barrels in the middle of the street as obstacles", "swap the third house for a tavern with a wooden sign".
GLB exports to Maya. Maya pre-vis artist takes over for camera animation, character animation, timing.
# What Yugma owns vs what Maya owns
| Phase | Yugma | Maya / Houdini |
|---|---|---|
| Initial scene blocking | ✓ | |
| Material direction | ✓ | |
| Lighting mood | ✓ (rough) | ✓ (final) |
| Camera animation | ✓ | |
| Character animation | ✓ | |
| Effects (smoke, fire, water) | ✓ | |
| Final cleanup | ✓ |
Yugma is the front-end of the pre-vis pipeline; Maya/Houdini still own the depth.
# The cost case
Pre-vis budget for a feature film: typically 1-3% of production budget, or $200k-$1M+. Front-loading the cheap 25% with AI saves cost without reducing quality — Maya artists spend their time on craft instead of asset placement.
# The studios using it
We have early-stage adoption from indie filmmakers and one mid-sized animation studio. None of them are replacing Maya. They're using Yugma to save the first week of every sequence.
# Honest limits
- Camera animation: not Yugma's lane.
- Character rigs: not Yugma's lane.
- VFX simulation: definitely not Yugma's lane.
- Final-quality rendering: Yugma is real-time WebGL; pre-vis final render is Maya/Arnold.