Learn to use "Blender"

[mtr]

https://x.com/wtry1102/status/2070046624374788277

I tried a workflow where I installed BlenderMCP, created a simple previz, and passed it to Seedance2 as a reference video.

The top row is a rough previz made in Blender.
The bottom row is the shot generated in Seedance2 using that as a reference.

I'm a Blender beginner (in fact, I touched it for the first time today), but by using Claude + Blender MCP, I was able to create a rough draft of the camera, composition, and subject positioning in about an hour while chatting.
I gave Claude a rough instruction like "make a scene with a mannequin holding a guitar in a space like a light music club room."

Then I had it add camera work, render, and output.

I use ChatGPT to craft a prompt that references that video.

I pass the key image made in Midjourney and the reference video to Seedance2 to create.

Being able to pass composition, zoom in/out, and camera movement—which are hard to control with prompts alone—to Seedance2 via video reference is a huge advantage.

The source is completely done with boxes, but since I could achieve something similar with a simple model that's easy to describe in natural language,
you can use Blender as an "AI video camera design tool."
Thanks to MCP, it's pretty easy to create.

This seems really useful for MV production.

[mtr]

https://x.com/reidhannaford/status/2070145120658137385

You can choreograph a complex action scene in Blender with basic shapes, then let Seedance make it real.

You need to try this AI filmmaking workflow:

1. Generate a start frame in Midjourney

2. Block out the action in Blender

3. Feed both to Seedance

My Blender reference was just rough timing, camera shake, and spatial choreography, and Seedance tracked the speed, motion, and action way better than I expected.

This is the difference between describing a shot and directing one.

[joe]

[mtr]

https://x.com/cheerselflin/status/2071976215632904342

https://github.com/cheercheung/Awesome- ... w-Usecases

[joe]