Tuesday, November 08, 2011

Visual Effects Details for Transformers: Dark of the Moon

Scott Farrar, Visual Effects Supervisor on Transformers: Dark of the Moon, continues to make the rounds in the Europe market to promote the upcoming November 28th release of the movie there. In the latest from WhatCulture.com, he continues to talk about the challenges of creating visual effects for Transformers. Below are highlights with the full article here.

Highlights:
- "[VFX team] are almost first in and last out when it comes to the movie" due to providing info on what will need visual effects and impact that will have on the budget.
- "Each character takes 15 weeks to create" as in just design the character with 15 layers of data to give it depth and texture (color, rust, etc). Example: "Optimus Prime has 10,108 pieces attached to his skeleton, that's over 150,000 levels of data just for Optimus."
- TF1 has 12 robots with 460 shots needing VFX; TF2 has 34 new robots with 550 shots and IMAX elements; TF3 24 new robots with 3D complicating things (not # of shots).
- Colossus (giant Decepticon worm) was "the largest asset that CGI has ever guilt, so large in fact it was close to shutting our computer system down." Our in this case being ILM. Two and half times bigger than Devastator with 86,823 pieces, 68 miles long resulting in 1.3 million dates pieces.
- Part of pre-production prep was filming and photographing the areas in Chicago that would be used for the film when recreating the city's destruction in the computer.
- For first time in 10 years Chicago opened all their bridges so production could use a helicopter to film the river route that the Tomahawk Cruise Missiles take in the movie.
- The voice actors play a part in how the robots are animated with facial movements and the like. Sentinel Prime, "with 100's of [face] pieces of metal moving", was pitched as being voiced by Sean Connery and animated along those lines. The casting of Nimoy, with similar facial structure, meant a few changes were needed to match his voice work.
- 350 people from ILM worked on DOTM by the end.
- Reason the car form is clean and shiny while the robot form is often dirty, damaged and rusty is because they discovered that Transformers looked fake when tried to match that new shiny metal look. Adding the dirt layer by later in the computer create that extra realism that sells the scene. (Too bad the movie extras never showed the "clean" look vs. "dirt" look to drive this home).
- Ferrar's favorite Transformers is Bumblebee: "he’s the guy who taught us how to do this movie………to try and show emotion on the face of a character who does not speak, it’s just like silent hero comedy, or tragedy. So the window of the soul is through the eyes and that’s absolutely true even in a robot"

3 comments:

  1. Martinus Prime11/09/2011 7:13 AM

    Whow, that shows how much stronger the computers are than a few years ago.
    I remember when they had to make Devastator in ROTF, there went a computer up in smoke.

    On one side I think it's a bit of a shame they used so much time and work and cgi detail power on a bot like Colossus. They could've better spend that on 2 combiners. But then again, we wouldn't have the awesome tilting building.

    Nontheless, awesome stuff from ILM!

    ReplyDelete
  2. will there be another transformers movie?

    ReplyDelete
  3. Do you know where this DOTM scale chart comes from? Is there a higher quality version of it out there somewhere?

    ReplyDelete

 
          Creative Commons License