In our most recent Scanning Stories post, we talked about making your full-body Skanect Structure Sensor 3D scans better using MeshLab and Meshmixer. In this fourth entry in the series, we’ll show you four steps between a full-body scan and printing that you’ll want to take to make sure your 3D selfies are true-to-life.
Step 1: Exporting your scans
Today, we’ll start by taking you through the process of exporting your Skanect scans as OBJ files. OBJ is a file format that contains 3D coordinates (polygon lines and points), texture maps, and other object info. By exporting your scans as OBJ files, you will get both a texture file AND a mesh file, as well as a file containing all data. This will help you better edit specific parts of your scan’s shape and appearance.
Start by exporting your Skanect scan file (for more info on creating scans, see this past post and this tutorial by Skanect) as an OBJ file.
Once you’ve exported the OBJ file, you’ll notice that Skanect has made three files for you: an OBJ, an MTL, and a PNG.
The MTL file contains all data, the OBJ is the actual mesh (the polyhedral version of your scanned object), and the PNG is the texture map, or surface detail file.
As a pro tip, the colors in the texture file tend to be a little too dark once printed so we we’ll show you how to lighten it up.
You can do this in programs like Photoshop or Lightroom.
Step 2: Compare textures before and after editing
If you look below, these textures look incomprehensible, but don’t worry, the computer understands how to read them. These are the colors that are what the computer is referencing to give texture to the print. The only problem is that the colors in the scan might not be vibrant enough once put onto the 3D model. Just like any form of photography we may need to do some image manipulation to make the colors look the best possible. Therefore, we’re going to bring this texture map into Lightroom and make some adjustments so they’re brighter and will look better.
The textures before lightning up
After lightning up
While editing the images, keep an eye on the details — they can be sharpened if needed. The settings we use in Lightroom are below:
| exposure +1.4 |
| contrast +40 |
| highlights +20 |
| shadows +10 |
Step 3: Editing the mesh and texture in ZBrush
When you scan with a hand scanner (like the Structure Sensor from Occipital that we use), you sometimes end up with a file that is not as sharp, complete, or accurate as you would like. You might have holes in the model that shouldn’t be there, or the texture might have flaws that need to be edited. You can make files like this printable using ZBrush or any other 3D program that can handle 3D files with textures like 3DMax, Blender, etc.
Here’s a super helpful tutorial for editing scans in ZBrush:
Zbrush tutorial to repair 3d scanned models
Other tutorials that might be helpful to you:
Pixologic ZBrush YouTube channel
Blender tutorial to repair 3D scanned models
How to Make Your 3D Scanned Models Look Amazing – Beginner Blender Tutorial
Step 4: Replacing the base of the scan
When you create a scan of a person, cleaning up the scan can often mean adding a platform, or base, on which the 3D print will stand. We generally remove the original base from the scan (which is either the ground, the floor, or a temporary platform the person was on) and replace it with a nice, freshly modeled platform because it will look much cleaner and stand upright.
To do this, follow the steps below:
1. Remember to always export the edited file in ZBrush as a VRML file. You’ll have something like this when you’re finished in ZBrush:
2. Make a platform. You can make a platform in any 3D modeling program. We made this simple platform in Solidworks. We made ours by drawing a square, extruding it to have thickness and then filleting the edges to make them less sharp. Once we were pleased with the platform we exported it as an STL file:
We then use Netfabb to merge the two files together. For us this is handy because our printers are set to work with Netfabb. Netfabb has a free version for you to experiment with.
3. Import the mesh AND your platform into Netfabb. Scale your model and/or your platform to a desired height/width. Place the platform underneath the model file (be sure they overlap, because if they don’t the finished file will end up as two separate parts and won’t merge during printing). Select both files and merge them together. Export the final file as a VRML.
4. Finally, create a zip file containing both the texture map PNG and the VRML model file and upload this to Shapeways.com.
The result is always a perfect finished product that can stand on its own.
Happy scanning and editing!
Brigitte & Astrid