Method Studios uses Enwaii LIDAR workflow for The Last Stand
The Last Stand is the first film with Arnold Schwarzenegger as lead actor since Terminator 3 Rise of the Machines shot 10 years ago. In this movie the international star embodies a sheriff of a small town close to the mexican border who faces a dangerous drug lord and his men in order to stop them to escape to Mexico.
The climax sequence of the movie is a fight between the sheriff and the drug lord. This fight takes place on a temporary brige installed by the gangsters over a long and large canyon in order to reach Mexico.
Only a small section of the bridge where the actors were fighting was built for real. The shot took place indoor surrounded with green screens. Method Studios was in charge of transforming the shot into an exterior shot by building a CG version of the whole bridge, the canyon and all the rest of the landscape.
In order to do so and to respect the tight schedule they faced, the work was divided in two teams : one in Method Studios London took care of creating the necessary assets, the other in Method Studios Los Angeles was responsible of delivering the shot.
Here are some shots of the final sequence which highlight the 3D environment recreated by the artists.
Images courtesy of Lionsgate and Method Studios
One kilometer of canyon
The canyon was approximately one kilometer long and required a lot of photos in order to be mapped with details. Ian Ward, CG supervisor of the Method Studios London team, opted for HDR panoramas. He requested that the photos were taken with GPS location to get a first approximation of the location of the photos. It helped to sort the photos because all of them were looking very similar ... rocks, rocks, rocks.
Ian decided to use the LIDAR calibration system of the Enwasculpt module of Enwaii to calibrate the photos accordingly to the LIDAR scans data of the canyon. Method Studios had already used this workflow successfully with other LIDAR scans of rocks for the movie Wrath of The Titans.
After the stitching process, the artists selected a dozen of 12K to 16K panoramas to cover the whole kilometer of canyon on both sides. They were able to manipulate such big images directly with Maya image planes thanks to the proxy system of Enwaii. The proxy system enables to specify a higher big depth and higher resolution image in order to switch to it automatically for texture generation. For instance, you can use a JPEG or TIFF image for the calibration and modeling tasks and texture with a HDR image in OpenEXR format when clicking the texture baking button.
One of the panoramas used by Method Studios (it is possible to zoom with the mouse wheel) :
During the calibration process, Peter Forsyth noticed that the LIDAR calibration was not as efficient with the canyon than it was with the Wrath of The Titans set, so he reported it to the support at Banzai Pipeline Ltd. It was indeed hard to identify specific features in the LIDAR scan since everything was so similar, almost fractal. So Enwaii developers worked on coding a second LIDAR calibration system, which is markerless and more adapted to calibrate organic shapes. This was satisfying for Peter and both systems are now part of the public version of Enwaii too.
Most of the shots of the fight sequence are taken from the middle of the bridge. Because there is no bridge in the real location it was of course impossible to shoot a 360 panorama from this place, which would have been convenient for mapping. So the artists decided to bake the texture of the 3D asset obtained with Enwamap without retouching. They rendered 6 images to make a 360 environment cube from the middle of the bridge and polished the environment using the rendered images. It proved pretty efficient this way. The surrounding desert was created using a Houdini setup blending with the canyon borders. Once ready, the assets were in the good hands of Method Studios Los Angeles artists who took care of delivering an impressive fight sequence.
We would like to thank Lionsgate and Method Studios for the material provided.