Tips on rigging up 3D scans for animation


(Punkoffice) #1

I recently uploaded a scanned model in an A-pose which you can see here:

Jessica Paige - A-pose by punkoffice on Sketchfab

This is not a sexy pose by any means, but the reason I scan in the A-pose is for animation rigging. Yes, a T-pose is preferable, but with camera rigs that can be too hard to capture. The hands are too far from the body. You would need to re-position the cameras to focus on the hands then you would have focus problems with trying to get the hand and body in the same shot. With an A-pose you can scan without changing the camera configuration. You can easily adjust rigs from Daz3D or Second Life to an A-pose or (as you can see in the video) you can quickly create a skeleton in Blender which you can use to pose your model into a T-pose so you can upload to auto-riggers such as Mixamo.

Here is a short video demonstrating the process:


(Bart) #2

Hey @punkoffice, thanks for sharing!

Did you do any weight painting to tweak the influence of your mesh?


(Punkoffice) #3

The weight painting was automatically done by the Mixamo auto-rigger (www.mixamo.com) Because its automatically calculated its not very accurate and you cannot manually adjust anything with Mixamo. If I rig an avatar for something like Second Life I will put a lot of effort in the weight painting.

Here's an example of rigging an avatar for Second Life. You can see the bits where I work on the weight painting (the colour gradients on the body-parts in Blender). Also, I scanned these guys in doing a T-pose and the fingers were so blurred cos the cameras couldn't focus on them properly. I lost all finger detail and they just ended up being some kind of melted hand :/