Completing the Moho Pipeline: From "Origami" 3D Environments to Multi-Angle Character Automation

Have you come up with a good Moho trick? Need help solving an animation problem? Come on in.

Moderators: Víctor Paredes, Belgarath, slowtiger

Post Reply
takuya
Posts: 46
Joined: Sun May 01, 2022 10:26 pm

Completing the Moho Pipeline: From "Origami" 3D Environments to Multi-Angle Character Automation

Post by takuya »

Image
Image
*(Caption: Instant 3D origami folding inside Moho using FoldWeaver 3D)*

Hi everyone,

I’ve been heavily invested in building custom workflow tools to streamline 2D/3D animation pipelines, and I wanted to share my development journey and the solutions I've come up with.
First, as a thank you to this great community, I have made two of my core pipeline tools—**RigWeaver** and **LayerWeaver**—completely **FREE** for anyone to download and use. You can grab them from my Gumroad store here:
https://2dpipelinetools.gumroad.com/

While developing these, I hit a massive wall: **Building matching 3D environments, rooms, and props directly inside Moho is incredibly painful and mathematically complex.**

Initially, I wrote a script with over 1,000 lines of code attempting to calculate target-snapping, nested parent group matrices, and Euler rotations to build 3D props. It was a disaster. Floating-point precision drifts and the sheer complexity of clicking targets destroyed any creative flow.

This led to a radical Copernican shift in my approach: **What if we discard the complex snapping entirely? What if we just snap selected edges flush to the Center/Origin Axis first, and then fold the layers using Moho's native Transform tools—exactly like folding origami?**

This concept eliminated coordinate drift and made 3D room/prop construction blisteringly fast. I packaged this "Origami" workflow into a new tool called **FoldWeaver 3D**.

But building the 3D space is only half the battle. In 2D animation, the most tedious and grueling task is **animating characters in extreme perspectives (high-angles/bird's-eye and low-angles/worm's-eye)** to match those dynamic 3D environments.

Traditional Moho workflows often fail here: either you can build a 3D background but the 2D character’s perspective breaks, or you force perspective on the character but it doesn't match the background.

This is exactly why the Weaver ecosystem exists as a unified, interconnected pipeline:

1. **The Environment (FoldWeaver 3D)**: Build your 3D rooms and backgrounds in extreme high/low angles using the intuitive origami method.

2. **The Character Perspective Engine (MotionWeaver + BatchWeaver + RigWeaver)**: they form a single automated relay system. **MotionWeaver** and **BatchWeaver** work as a multi-threaded hub to translate and batch-export your 3D motion data (BVH) into multi-angle 2D coordinates (front, side, high-angle, low-angle). Then, **RigWeaver** instantly takes that exact angle data inside Moho to align, scale, and synchronize your 2D character to perfectly match the exact perspective of your FoldWeaver environment. No more characters looking flat inside a dynamic 3D room.

3. **The Surface Details (MeshWeaver & PinWeaver)**: Two direct-projection engines that bake 3D motion perfectly into camera-facing 2D smart warps and pin bones. This allows you to draw cell-shaded shadows or flowing accessories directly on a flat 2D canvas while they stick perfectly to 3D rotating props.

Check out the full workflow in action

I wrote a highly detailed blog post about this exact development journey, explaining how this entire "Weaver" ecosystem functions together to eliminate coordinate drift and automate perspective generation:
https://sentakuclip.net/blog/8151/

To make this entire workflow accessible, I’ve released **FoldWeaver 3D, MeshWeaver, and PinWeaver** today on my Gumroad for just $5 each (basically the price of a cup of coffee). For those needing the top-tier motion translation and batch-generation pipeline, **MotionWeaver** ($20) and **BatchWeaver** are also available.

I would love to hear from other animators and scripters here:
- How do you currently tackle 3D sets and extreme perspective matching directly inside Moho?
- Have you ever fallen into the "coordinate drift" trap when nesting 3D layers?

Please feel free to download the free tools, read the dev story, and let me know your thoughts!

Best regards,
takuya
Posts: 46
Joined: Sun May 01, 2022 10:26 pm

Re: Completing the Moho Pipeline: From "Origami" 3D Environments to Multi-Angle Character Automation

Post by takuya »

Just wanted to add a quick thought on the core philosophy behind the Weaver series, in case the technical details made it sound a bit overwhelming!

Ultimately, this entire ecosystem is designed to land perfectly in 2D. The 3D space isn't the main character here; it’s simply an "invisible scaffold" to help us draw beautiful hand-drawn lines, cell-shaded shadows, and flowing accessories in perfect perspective.

If you are familiar with illustration tools like Mental Canvas or Feather, the concept is actually very similar.
Mental Canvas creates a sense of depth not by 3D modeling, but by placing flat 2D canvases in space like papercraft—which is exactly the inspiration behind FoldWeaver 3D. Feather allows artists to sketch freely in space without typing in complex Z-axis numbers.

I wanted to bring that exact intuitive, "illustrator-first" feeling into Moho:
- RigWeaver translates 3D spatial motion directly into 2D animation controllers.
- MeshWeaver and PinWeaver extract 3D coordinates and bake them onto a camera-facing 2D screen plane.
- This means you can draw, paint, and adjust vector lines just like drawing on a flat piece of paper, completely ignoring 3D coordinates or UV unwrapping.

As riggers, it's easy for us to get trapped in the mechanical mindset of bone structures and mesh layouts. But the true goal of these tools is to step away from the technical constraints and treat Moho's 3D space as a free, intuitive canvas for 2D artists.

Has anyone else here experimented with this kind of "spatial drawing" approach instead of traditional rigging? I'd love to hear your thoughts!
takuya
Posts: 46
Joined: Sun May 01, 2022 10:26 pm

Re: Completing the Moho Pipeline: From "Origami" 3D Environments to Multi-Angle Character Automation

Post by takuya »

Hi everyone,

Following up on our discussion about treating 3D as an "invisible scaffold" for 2D animation, I’m excited to share a major update to **MeshWeaver** along with a comprehensive demonstration video!



manual
https://sentakuclip.net/jetb/wp-content ... Manual.pdf

In this video, I demonstrate the newly updated MeshWeaver workflow by building a stylized 2.5D skyscraper with full 3D rotation and action synchronization.

Here is what's new in this version of **MeshWeaver**:

1. **3-Point & 4-Point Surface Projection Support**
In addition to standard 4-point quads, MeshWeaver now fully supports 3-point triangular surfaces. This makes it much easier to project flat vector canvases onto triangular facets, sloped roofs, and complex polygonal meshes.

2. **Cross-Layer Point Selection**
You no longer need to merge or reorganize your vector artwork beforehand. You can now freely select guide vertices across multiple separate layers to define your target projection surface.

3. **Instant Action & Timeline Synchronization**
This is the biggest workflow breakthrough shown in the video: once you animate your 3D layer (such as a continuous turntable rotation), running MeshWeaver's **Update** instantly bakes and synchronizes the action keyframes across all rigs!
Even if you manually tweak or deform individual points on the canvas (like distorting specific windows for custom artistic styling), MeshWeaver preserves your hand-drawn edits non-destructively while maintaining flawless synchronization with the 3D motion.

**A Quick Note on the Asset Workflow (CorelDRAW Text-to-Vector):**
In this tutorial, I also experimented with CorelDRAW 2026's new **Text-to-Vector AI** to generate the skyscraper’s detailed window facade. Generating the base vector grid from text prompts, cleaning up the SVG paths, and bringing it into Moho proved to be a rapid way to create rich architectural assets. Combining AI vector generation with MeshWeaver’s 2.5D surface tracking opens up a really exciting avenue for background artists who want quick, complex vector assets that move seamlessly in perspective.

As always, the updated tools are available on Gumroad (https://2dpipelinetools.gumroad.com/).

I’d love to know what you think of this action-sync approach, and how you currently handle detailed textures and architectural facades across rotating 3D setups in Moho!

https://sentakuclip.net/blog/8181/

Best regards,
takuya
Post Reply