News

The World in 3D

As a 3D visual effects artist, I spend a great deal of time recreating and manipulating reality. Sometimes it is rotoscoping a person out of video footage and erasing any trace of that person. Other times it involves creating a 3D character that is animated and inserted into the footage. And other times an actor is inserted into a completely digital world.

During much of this, there is a point when the flat image of the footage must be recreated in a 3D world. This can be done using tracking software most of the time, but still takes a great deal of dedication.

Standford has come up with a technology that could really help the digital artist in years to come as this technology matures and becomes commercially available.

At the most basic level, they have developed a camera that can take photos in 3D.

Here is how this new camera operates. Normally, a digital camera has one sensor behind the lens. This sensor sends the data to a processing chip to turn that data into an image. A crude version of the human eye.

The Stanford camera is designed more like a fly’s eye. In lieu of one eye, the sensor is comprised of a 16×16 grid of smaller eyes. Software then compares each side by side sensor and computes what has changed. Then it calculates how far away each element is from the camera based on the movement. Large change from one sensor to the next means large distance.

How will this affect the digital artist?

With depth information, you could easily pass an object or another person behind someone else already in the footage. This can be done today, but involves tedious rotoscoping. With this tech, it would be as easy as stacking the ‘layers’ in a compositing program like Shake (what we use as our studio’s compositor of choice).

Here’s the link to the article:
http://www.news.com/8301-13580_3-9874436-39.html

© 2008 Reels in Motion Visual Effects News.