Blender – somenice

Generative 3D Modeling

Recently Microsoft released the open 3D generation software Trellis.
I decided to try a few tests as reviews had been quite favourable with results creating 3D models with “AI”.

The website describes the technology as unified Structured LATent (SLAT) representation and there is a demo page to try it out.
https://trellis3d.github.io/

I chose an image-to-3D example, meaning I upload an image and get back a textured 3D model. Having been a long-time fan of photogrammetry, or creating geometry from two dimentional images, and also a 3D genralist that regularly creates 3D models using Blender, I was eager to see how it would perform.

Using the prompt example seen on VFX & Chill live stream where two artists used Z-Brush to sculpt food in a 20 minute timed challenge, I set out to create a seasonal gingerbread model.

First, Adobe’s Firefly image generation created this image using the prompt: “elaborate gingerbread house with icing“.
Impressive start.

Machine generated image using Adobe Firefly prompt "elaborate gingerbread house with icing"

Next, dragging the image to the demo site and clicking “Generate” with default settings impressively separated the image from the background and in about 30 seconds showed what it thought the full model would look like. The entire unseen back of the house is amazingly “imagined”.

What’s produced is a GLB file that contains mesh and texture that can be dragged into your 3D program and edited or used as any other asset. In about 4 minutes I have any model loaded ready to use.

As for the model quality, it created a 52K face mesh count which could be considered medium quality. There are lots of artifacts and non-manifold holes but for any background shot it might be just enough to pass the test. Some retopology is almost certainly required and honestly given the choice to “fix” this model or start from scratch I might be tempted to just build it again.

Great from far, far from great.

I don’t give it long, maybe two years, before anyone can drop any fully 3D asset they can think of into any scene realtime.

As with any AI/ML generative model please understand the environmental and ethical implications.

Blender Title Animation (Sound On)

80’s Logo inspired by PBS WGBH Boston Station Identification

Cloth Simulation with Procedural Plaid

Procedural cloth with plaid material. Made with Blender

This was fun, because how can you have too much plaid?
Thanks to Ryan King Art on Youtube for the Blender tutorial series on material nodes. His tutorials are great and I appreciate the effort in his editing. Support him if you can.

I often joke that I’ve learned Blender 3D every year for the last 20 years. Sometimes because I forgot what I previously learned and don’t use it frequently but more often it’s because a new release offers new tools and capabilities.
The development has seen major changes to the UI and entire interface in recent years.
The latest have been geometry and material or shader nodes. A way to procedurally or programatically link blocks, each affecting the output in a chain. Either to create geometry or to produce any desired material. Plaid in this case.
A previous post shows another example using nodes to create and animate fake trees.

It can get messy and may look confusing but when grouped the key attributes are easily accessed and changed. It makes sense that the Blender Add-On called “Node Wrangler” comes in handy.

Blender is a free and open source 3D application, completely extendable using Python. It continually delivers updated features and is capable of producing amazing compositing/animation/video/3D printing, you name it. (I do miss the game engine)

If you’re looking to learn 3D I highly recommend Blender 3D
And again look for Ryan King Art on Youtube for plenty of beginner tutorials.

Procedural “Trees” using Blender Geometry Nodes

I’ve been having a lot of fun playing with Blender Geometry and Shader Nodes.
Specifically here using Instances on Points. I started with a 60 x 60 grid of curves, made into spirals and animated with a Noise texture.

This ended up creating 8.3 Million faces and took 12 hours to render.