3D Week 8: 3D scanning

Hello there! During this week we worked with the modeling of human faces. We learned about different methods of creating 3D meshes of faces. We also talked about the topology and edge-flow for facial projects.

The method we used was to convert real life objects into digital 3D meshes. We did this by scanning photographs of the real life object from many different angles. The photos had to cover the entire face, otherwise there would be holes and missing parts in the digital 3D mesh resulting in a bad model.We were given the opportunity to work with our own faces, to create digital 3D meshes of our own faces. The amount of photos you take determine the quality of the scan, the software will have more information to match the photos and create a 3D model from them.

Once all of the photos are taken you upload them into your computer. We used Autodesk 123D to create the digital 3D mesh from the photographs, it is simple and free. When that is done you can download the model as an .obj, so that you can open it in your 3d software.

This was the first part of the of the exercise.

In the second part of the exercise we were tasked with changing the topography of the digital 3D mesh. The edge-flow is very important if you want the character to be able to make expressions, and move its face. The edge flow on the face was rather simple. The edge flow goes around in circles around the mouth and eye areas. These are also the most important parts of the edge flow when working on a face. Modeling half of the face can save you a lot of time, but you have to watch out, almost no one has a completely symmetric face so making slight changes to one side could be a good idea.

We used a trial version of the program 3D-coat to work on the digital 3D-mesh.

We were also supposed to read the article  ‘Photogrammetry’ in ‘The Vanishing of Ethan Carter’. I really enjoyed reading the article. I think that the author explained the difficulty in a rather informative way. I have never tried Photogrammertry before, well not “real” photogrammetry, so at first I kind of just thought “well it’s probably difficult and requires skill to get a really good 3D scan out of it. But how hard can it really be?” After continuing to read the article I started to understand the work and effort which you have to put into photogrammetry. And it is really a skill which you can master. Even when us students were just going to take pictures of our faces a lot of the models did not turn out well because they were blurry. It kind of puts it into perspective when we were just supposed to take around 10 pictures and barely made good models, and then when the guys in the blog are working on taking around 250 photographs so that they can get an entire house 3D modeled.

Something which I thought was very interesting about the article, was in the end when he was writing about not wanting to preserve the photo realism to a hundred procent. They did this to support a certain visual style which would fit in with their game.