Recently I worked on a new 3D scanning project using the new Revopoint POP 4, and this time I decided to give space to my nerd side by testing it on an iconic anime character: Ken Kaneki from Tokyo Ghoul.
The action figure measures 203 × 177 × 185 mm and, due to the complexity of the subject, I decided to approach the scan using multiple scanning sessions and different acquisition methods depending on the part being scanned, assembling everything afterward inside MeshMixer.
For the main body of the figure, I used the marker frame together with the Global Marker tracking mode. The scan was performed using Cross Lines laser mode, High Accuracy, and a point distance of 0.15 mm. The point cloud was fused at 0.13 mm and, once completed, the frame was quickly removed, leaving only the main subject.
For the two Kagune parts (the “wings” coming out from the character’s back) and the base, I used a different workflow. Each part was positioned on small supports placed on the turntable and scanned in two separate sessions by rotating the object 180 degrees. Again, I used Cross Lines laser mode, High Accuracy, and a point distance of 0.15 mm. Once the two point clouds for each object were completed, they were merged together directly inside RevoScan 6 using the Merge function.
The head was scanned separately using Full Field mode, with the object placed on the turntable while the scanner was mounted on the articulated arm. The turntable rotation speed was set to 200 seconds per revolution in order to capture as much detail as possible from every angle.
Once all the scans were completed, I generated the meshes separately using a mesh resolution of 0.15 mm. All the parts were then imported into MeshMixer, aligned manually, and assembled together using the Merge command.
The final result honestly surprised me. Every small detail was captured extremely well, from the anatomical parts and clothing folds, to the organic shapes of the Kagune and the stone textures of the base. Even the fine surface details on the Kagune were reproduced very accurately.
This project was particularly fun because, although it was definitely not the simplest scan to execute, the workflow itself was much easier and more stable than I initially expected. Using different scanning approaches for different parts of the same subject made the whole process much more manageable and allowed me to capture areas that would have been difficult to scan in a single session.
As always, no scanning spray or surface treatment was applied before scanning.
What do you think about the result? Have you ever used multiple scanning modes and workflows for a single object?
POP 4 Kickstarter: