3D scanning is the non-invasive process of replicating any existing person and creating its virtual representation. The end product is a photorealistic 3D CG model. While these models are most commonly used to support VFX work in film, television, and video game production, the possibilities are endless. The process is simple, fast, and efficient.
There are a number of situations CG characters can be utilized on screen. Any time you want to have visual effects affecting a human character it's good that have that character in 3D. This gives you a ton of options to play with, and in the 3D world which allow for shape shifting, stretching or any type of deformations as well as particles exploding. Another use is for stunts or death scenes in which you do not want to subject the actual actor to any harm. The digital double mimics the likeliness and can be the 'pin-cushion'.
If an actor is unavailable for any reason, a digital double could be used in that scene. Blend shapes are fascinating in that if you retain the same 3D mesh, poly count, and vertex order, you can essentically 'shape shift' from one face to another face. This can be used to create facial expressions for animation.
For example we always scan a neutral expression head first, and if we did scan a smile, once the blend shapes are created, you can with a slider go from a neutral face to a smile, and all the varying degrees between the two. If you were to scan 12 different people's heads and created blend shapes for them, you could mix and match features to create new characters - for example 20% of person 1, 30% of person 3, 50% of person 9. The result would be a unique character, and interestingly enough it would not be entirely fiction, but accurate human features.
The first step we do is to take some initial photography shots of the person exactly as they are. This will allow them to relax as well as let us capture their hair exactly how it was. We shoot a full rotation of the person to have full coverage of high resolution photography for both reference to ensure the 3D scan matches the actual person as well as for later on projection mapping the textures.
During the actual head scan, the person being scanned is to remain somewhat still for about 20 seconds. This is because when scanning static objects the shape can never move, and so recording data is easier in comparison. A human on the other hand, and especially the head, will be moving as they are alive and breathing. We only ask they do not move or lick their lips for the duration of the scan. This lets the volume box we're capturing remain constant resulting in better alignments and ultimately better resolution in the final product.
We have successfully scanned animals as well as a human baby, so by no means does someone have to be perfectly still. We appreciate the person's cooperation.
A human body is harder to scan then a head, but is entirely possible to achieve high resolution data. The first reason it is more difficult is because of the surface area. In relation to a body, there could be a dozen heads to account for the surface area. What can easily end up happening are misalignments due to a person moving their legs separately from moving their arms, and their arms separately from their torso, and so on.
This localized movement can be trouble for some types of alignment which try to use large chunks of scan data at once. We are able to compensate for all localized movements in our pipeline with our researched techniques and experience. The photographic reference is more elaborate than just for a head as there are 4 limbs, hands, feet, torso, and hips. The more time allotted to photography will generally improve the color texture in both resolution and color information.
For example if you framed an entire body with your camera versus framing just the upper body, you'd get more resolution and detail out of the latter.
"What can I say about Nick Tesi except that he always delivers and does a terrific job. Over the past many years, I have had great experiences with Nick Tesi when I hire him to do any type of cyber scanning. His 3D digital models are superb and are of very high quality. I know I can count on Nick to help create digital actors or digital models for any type of show. I would highly recommend Nick and his team for projects that requires any type of 3D cyber scanning. Thanks Nick for your help on "Vamps." See you on the next big one!"
Susan Zwerman - VFX Producer“VAMPS”
"I selected TNG Visual Effects and Nick Tesi for my 3D scanning needs based on our past history on other projects. He has always delivered on time at the price we agreed upon. His company is reliable and always proved to be a good choice of vendors. On my last project Nick arranged Lidar scanning and all worked out perfect. Nick has always been prompt, and helpful in supporting me to find the correct solutions."
John Sullivan - VFX Supervisor
TNG Visual Effects copyright 2014 | Ph: 877.879.2040 toll free | 3424 W Carson St. Suite 210, Torrance, CA 90503