It Started with a Portrait

This project began with one simple question:

“Can I turn my own cartoon-style portrait into a real wooden sculpture that I can hold in my hand?”

I started with a personal character image. Instead of keeping it as a digital image, I wanted to explore the complete process from a 2D portrait to a 3D model, CNC toolpaths, 4-axis machining, and finally a physical wooden object.

Caption:
The starting point of the project: my personal cartoon-style character image, used as the reference for the 3D model and CNC sculpture.

From a 2D Image to a 3D Model

I used Makera Studio and its 3D Generation tool to convert the character image into a 3D model, then exported it as an STL file for CNC machining.

When I first saw the flat image become a model that I could rotate and inspect in 3D, I thought the hardest part was already done.

Once I started machining, I quickly learned:

Creating a 3D model and successfully machining it are two very different things.

Caption:
Using Makera Studio to convert the character image into a 3D model and export an STL file for the next stage of physical fabrication.

The Real Challenges Started with Machining

After preparing the STL model, I began setting up the stock dimensions, rotary center, Z-zero, machining orientation, roughing toolpaths, and finishing toolpaths.

At first, I thought that once the tools and toolpaths were ready, I only needed to wait for the machine to finish the sculpture.

The real process was very different.

I encountered the tool touching the stock before reaching the intended cutting position, the first cut going too deep after probing, tool interference with the model, areas that were not machined as expected at certain rotary angles, and long periods of air cutting during finishing.

As a beginner in 4-axis CNC machining, every solved problem seemed to reveal another problem waiting behind it.

One Interrupted Job Created a Face on Both Sides

One of the most memorable failures happened after a machining interruption.

When I restarted the job, the A-axis rotary position was no longer correctly aligned.

As a result, facial features began appearing in an area that should have been the back of the character.

That failure taught me an important lesson:

In 4-axis machining, the A-axis angle is part of the machining coordinate system too.

Recovering an interrupted job is not simply about finding the correct line of G-code. The rotary position, machining coordinates, and machine state must also be considered.

Caption:
The front of the character was taking shape, but the A-axis position became incorrect after restarting an interrupted machining job.

Caption:

The other side of the same failed attempt. Facial features were machined where the back of the character should have been, showing how important rotary positioning is when recovering an interrupted 4-axis job.

More Than Six Failed Attempts

This project did not succeed on the first try.

I went through more than six failed attempts, interruptions, setup changes, and machining tests.

Some attempts had positioning problems. Some toolpaths did not produce the result I expected. In other cases, I only discovered after machining had started that the model, support structure, or setup needed to be changed.

I kept several of these failed pieces.

Looking back, each failed piece represents a question that I did not know how to answer at the time.

Caption:
Several failed attempts from different stages of the project. None of them is perfect, but every failure removed one unknown from the next attempt.

A Beginner Has More to Solve Than Just the Toolpath

Through these attempts, I encountered many questions I had never considered when I started:

Why did the tool touch the stock before machining started?

Why did the first cut suddenly go too deep?

Why does the actual maximum stock diameter matter?

Why were some rotary angles not machined as expected?

Why could the tool tip reach the model while the tool body still caused interference?

Why was the finishing toolpath spending so much time cutting air?

What should I do when spindle temperature rises during a long machining job?

How can I safely continue after machining is interrupted?

These problems gradually changed my understanding of CNC machining.

4-axis CNC is not simply importing an STL file, generating G-code, and pressing Start.

The model, stock, cutting tool, workholding, Z-zero, rotary center, toolpath strategy, machine condition, dust, and heat can all influence the final result.

Building My Own Process Through Failure

After repeated adjustments and tests, I gradually developed a more complete workflow:

Personal Character Image
↓
3D Generation in Makera Studio
↓
STL Preparation
↓
4-Axis Setup
↓
Stock Rounding and Reference Setup
↓
3D Roughing
↓
Fine-Tool Finishing
↓
Checking Tool Interference and Unmachined Areas
↓
Toolpath and Parameter Adjustments
↓
Finished Wooden Sculpture

In the end, I did not discover one “perfect setting” that would prevent every failure.

What I learned was how to identify where to start looking when something goes wrong.

More Than a Finished Sculpture

This project originally began as an attempt to turn my own character portrait into a wooden figure.

After more than six failed attempts, it became something more.

The successful sculpture matters to me, but the failed pieces are also part of this project.

For an experienced CNC user, Z-zero, stock diameter, tool interference, A-axis positioning, or interrupted-job recovery may be familiar topics.

For a beginner entering the world of 4-axis CNC, each one can become a real obstacle that must be understood and solved.

That is why I want to share not only the successful result, but also the mistakes, troubleshooting, adjustments, and attempts that came before it.

If another beginner someday asks, “Why did my first cut go too deep?”, “Why was this area not machined?”, or “Why did my rotary position change after an interruption?”, perhaps something from my failed attempts can help.

Making is not always about succeeding on the first attempt.

Sometimes the most valuable result of six failed attempts is finally understanding why the seventh one works.