24 things we learned about Teenage Mutant Ninja Turtles by visiting ILM

The folks at ILM invited Matt to go and talk about Teenage Mutant Ninja Turtles with them. They didn't have to ask twice...

When Den of Geek visited special effects company ILM to talk about the new Teenage Mutant Ninja Turtles film, we did more than gawp at clips of the film and eat sandwiches. We also learned things. 24 things, to be specific.

These are the 24 things we learned.

1. The sandwiches at ILM are delicious.

2. According to SFX Supervisor Pablo Helman, ILM had an on-set footprint on Teenage Mutant Ninja Turtles production of “about fourteen people”.

Ad – content continues below

3. We learned how ILM breaks down a shot. Animation Director Tim Harrington said, “We send it into body tracking. We have three cameras around the set that triangulate the performance of the actor. We track that movement with a digital version of the actor. Once we have that, we take that motion data and we apply it to the Turtle.”

The first version of the animation is “kind of simple and blocky at this stage. It’s missing a lot of the detailed simulation. It’s missing the face, the fingers aren’t moving. It’s just the raw mo-cap thrown on the Turtle. Then, it goes through final animation, final sim, muscle simulation, final lighting and comp.”

Of the renders they showed us, the first was weird looking, the second was clumsy and gray and the third was utterly perfect.

4-9. Tim Harrington also gave us the following stats. We are counting each one as a thing we learned as each one is a thing and we learned it. The quote from Harrington goes;

“6 animated hero characters579 shots with the animated characters200 were close up dialogue speaking shotsHero characters are on screen for 48 minutesat our biggest, we had 60 animators181 individual facial controls for each Turtle, over half of those were just for the mouth.”

10. The highly detailed CG stuff is referred to as being ‘hero’. The way we learned this was by becoming confused. When they told us they had six hero characters, like in the stats bit above, they kept referring to the four Ninja Turtles, Splinter and Shredder. “Jesus wept. Shredder is a villain, guys!” we thankfully chose not to shout at them.

Ad – content continues below

Then, they started to refer to hero buildings in the CG cityscape they’d created and we cracked the case. Den of Geek publishes a lot of Sherlock content and we’ve apparently absorbed some his sleuthing skills.

11. The martial arts scenes feature the use of CG to exaggerate the physical performance of the stunt guys. 

I like to think of it as using CG to rewrite the word ‘punch’ with the caps lock on.

12. It’s really difficult to focus on an interview when they’re playing you a clip of the Ninja Turtles fighting Shredder. 

13. The Shredder costume and blade system are mechanically accurate. He’s like a Swiss Army Shredder.

14. Pre-CG Shredder looked like an old version of Shredder by coincidence. We got to see some pre-CG Shredder footage and the mixture of motion capture gear and blades gave us flashbacks of Shredder from the TV show Ninja Turtles: The Next Mutation. Thankfully the CG’ed version looks completely different, but it gave us a hell of a fright. The Next Mutation is not one of our favourite television programs.

Ad – content continues below

15. Raphael’s clothing is based on a gladiator costume. “Because basically he is a gladiator. He’s a warrior. He’s a fighter,” said Tim Harrington, who also said that the character was partly inspired by Clint Eastwood.

16. The Turtles designs were practical as well as artistic. Tim Harrington said, “The challenge was making them (the Turtles) talk. It’s actually really hard to make them talk with a turtle beak. They have to have mouths.”

17. Unfinished animation looks a bit rubbish. We have no idea how they can tell if it’s going well or not. Some of it looked blocky, other bits looked a bit like clumsy stop motion and some of it looked like terrifying clay monsters. At what point during the clay monster production phase are you able to say “Yep, this is what it’s meant to look like”? But this, alongside the absence of any relevant skills, might be why we don’t work at ILM.

On the same note, we were shown at one point how they’d made a blocky mess. Then, it became a digital version of the physical lair set. To our eyes, it was indistinguishable from the real thing. We found what ILM do to be absolutely awe-inspiring.

18. A lot of the backgrounds featured in the movie were created digitally. It was kind of a big deal. Associate VFX Supervisor Robert Weaver told us that “Just building the Time Square area probably took a team of 12-15 artists a span of 6 months. That is to set up the work for all of the other artists to come on and create the actual shot work.”

19. They were able to show us frames from the snow chase sequence that was entirely CG. The snow environment is very detailed, with each pine tree deliberately placed and several layers of snow on the ground. “It’s an environment that Jonathan (Liebesman) wanted there to be a tremendous sense of peril every step of the way.” Robert Weaver told us.

Ad – content continues below

20. High tech still leans on low tech. Associate Animation Director Kevin Martel keeps a mirror on his desk so he can refer to his own face. He doesn’t specifically look like a Ninja Turtle, one in particular or in general, but it’s just useful to see what certain facial expressions looks like at times.

21. CG eyes can look very, very real now. As per Robert Weaver, “Initially we put a tremendous amount of detail into the irises with refraction, with sub surface scattering. Just a lot of work goes into it to lend itself to that photo realism quality.”

They showed us an eye on a big screen that looked so much like a human eye I became concerned that they might one day make something that could physically pass for a human being. ILM worked on Terminator 2 and now they’re kicking out serious SkyNet vibes? Worrisome.

22. The data they capture from the MUSE system looks kind of wriggly.

To elaborate on that, they showed us what the data they capture looks like in a raw form. Now, I don’t know if they designed it to look like this as it’s a form they can understand or if they were placed in a Gozerian ‘choose the form of your data’ subconscious selection type of deal. But it looked like the output from a heart monitor, assuming the user was drinking idiotic quantities of Red Bull or perhaps watching Crank at the time of the reading.

The jagged line sits in the middle of the screen, which sets of sliders either side. The sliders control a muscle system within the CG face.

Ad – content continues below

23. ILM have a team who deal with gravity. We don’t mean they have a group of people at war with physics, of course, but rather a bunch of people whose job it is to make sure that CG elements interact with gravity in the same way that the real life things they’re simulating would. 

“So, he’s (Mikey) got all his chains and his glasses hanging down. We’ve got another department who just use this amazing simulation technology that gets gravity and weight just right. The way it collides with their bodies. It brings just an extra level of realism to it,” Kevin Martel explained to us as we nodded along.

‘A person who uses a computer to pretend there is gravity – of course’ we thought, and reasoned that we’d understand everything we were being told on the third or fourth listen back to our audio recordings. We’d officially reached the point where we couldn’t tell if the technology was incredibly advanced and we were thick or if it had been a long day full of techy talk and we just really needed a coffee.

24. The coffee shop at ILM is called Javva the Hutt.

Follow our Twitter feed for faster news and bad jokes right here. And be our Facebook chum here.