About X Men 3
“X-Men 3: The Last Stand,” follows the tale of our favourite groups of mutant super beings. In this film we see the effects of humans finding a, so called, cure for mutation and it has obvious repercussions which result in a war amongst mutants.
Official Movie Site:
About Framestore CFC
Framestore CFC is the largest visual effects and computer animation studio in Europe, with over 20 years of experience in digital film and video technology. The company has won numerous international awards including two Technical Academy Awards from the Academy of Motion Picture Arts and Sciences, three BAFTA Craft Awards and eleven Primetime Emmy Awards.
Framestore CFC Website:
VFXTalk Meets The Man Who Helped Put The X in X-Men 3
Welcome to our latest interview in the VFX Master series. We asked you to give us questions for Craig Lyn, VFX Supervisor representing Framestore CFC’s work on, ”X-Men 3: The Last Stand.” Once again to your credit, you gave us plenty of questions for Craig. We got to sit down with the man himself and get you what you wanted…answers.
Craig’s early career in computing led him to write a book entitled "The Macintosh 3DHandbook" (currently in its 3rd edition), and his writing about 3d and compositing for film led him into 3D visual effects. After freelancing for several years, Craig eventually joined ILM for Star Wars: Episode 1, on which he worked in a variety of capacities. At ILM he broadened the range of his interests, ending his tenure there as a Technical Director in the CG Department.
Craig joined Framestore-CFC as Lead Technical Director on Dinotopia (which won an Emmy and four VES Awards), subsequently sequence supervising on Harry Potter and the Chamber of Secrets. He was the CG Supervisor on Thunderbirds, for which Framestore CFC created over 600 VFX shots - one of the largest and most comprehensive undertakings ever achieved in the UK. Craig has just finished the work as VFX Supervisor on X-Men 3: The Last Stand.
Can you please tell us a little about yourself and your history in VFX?
CL: I came over to the U.K around 5 years ago to start working on, “Dinotopia,” which was a long format show for Framestore CFC. Before that I was at ILM in the bay area of California which is where I got my start in the VFX industry. After 4 years at ILM the last show that I worked on was, Michael Bay’s “Pearl Harbour” and really wanted a change, the U.K seemed like it was a market that was about to boom and so I came over here.
I was on the Emmy team for Dinotopia. I also worked on Mistress of Spices, a few of the Potter films, Star Wars Episode 1, Thunderbirds, Mission to Mars, Galaxy Quest, Amistad, Deep Impact, Mrs. Henderson Presents, a little bit on Doom, and many commercials.
Can you just explain the process of what you did on the bridge scene?
CL: We were bidding against ILM for this sequence and both myself and Mei Ming Casino our VFX Producer, used to live in the bay area. Also I used to drive over the golden gate bridge all the time…so we had 2 great incentives for wanting to win the sequence.
Another company had started off doing the pre-vis, and there were a lot of changes going on, The show was constantly in flux it is very important for you to realize that…originally it was supposed to be Bryan Singer directing, then Matthew Vaughn and then finally we ended up with Brett Ratner, with changes like that the story is going to change. The pre-vis was done and then redone and we ended up doing a lot of it as well, in fact I did a lot of it on my laptop out in Vancouver.
The scene itself plays out in two main parts: The practical set piece and then the big CG shots. The set piece was built out on location in Vancouver BC and was 6 lanes across and around a hundred of metres long…that was used for all of the live action pieces on the bridge. The change of directors led to a change in DOP’s which meant changes in lighting strategies. At one point they just couldn’t decide at what time of the day that sequence was going to be…originally it was a night time theme, the lighting was set up for that and it would have worked…but things changed and they ended up wanting it to take place at golden hour, however, it was shot at night, in the rain on a green screen…so we ended up shooting night for day, which is one of the worst conditions you can have. We had to deal with spill, light bounce and so on. We fought with it all the way through, but I think it turned out pretty well. The second bit was all CG…the problem with that was the constant change in vis and plates. Since we often didn’t end up with the plates that we wanted, we had to create a lot of matte paintings and digital environments.
What did you use to do it with?
CL: The majority of the animation and lighting was done in Maya. Some of the rigid body simulations and all of the dust work was done in Houdini. The cable simulations were done in Sy-flex, which is a plugin for Maya. The compositing was done in Shake, rendering done through RenderMan and a lot of custom code for miscellaneous parts of our pipeline.
What kind of research were you able to do?
CL: Actually the Tacoma Narrows Bridge in Washington, would you believe that there was a suspension bridge that was built and the frequency of oscillation matched what the wind was doing, so when the wind was blowing…the bridge would wobble and eventually one day it flew apart. That was great reference footage. The great thing about the sequence is that everyone knows what the Golden Gate Bridge looks like, you can find reference for it everywhere you would look. We could find the lighting and anything we needed from any postcard.
What did it take to model, animate and composite the wings of Angel?
CL: It was all of our custom code. It’s legacy code, which is a good and a bad thing because we adopted another show’s work. We took the code from the Hippogriff, (which was like a Pegasus,) in the Harry Potter movie. That was part of the reason why we got the work. We used a lot of that code to create the wings for Angel. Parts of the wing was procedurally generated, specifically the smaller feathers.
How exactly did the experience form Harry Potter with the Hippogriff help you out there on the modelling and animation side?
CL: In terms of the bidding process it helped us a lot because we were able to offer an off the shelf solution. The problem with that of course you don’t develop something specifically for the movie.
How many feathers and muscle systems make up the wings?
CL: Feathers oh my, thousands, in fact hundreds of thousands because there are all the small little downy feathers, it’s not just the big ones. There wasn’t as much of a muscle system, as a dynamic system…you know getting the flap and the flutter and all of that. The muscle system was just clever rigging, trying to get something to fold up like that is very very difficult, in fact getting a wing to fold is a rigging nightmare to begin with.
What was it like building the skeleton/ik chain for the wings?
CL: A lot of that work was already done for us from another show. A wing really is a wing and we didn’t have to modify it that much. We did in some ways, but if you imagine any guy with wings on his back is going to look a little goofy. It was prosthesis in shots where the wings were folded up on the back of the actor, and that was done on set by another company. When the wings were spread and flapping that was us.
Did you use any off the shelf plugins?
CL: Sy flex we used for the cable sims, which worked very well and did it’s just, but that was it really for plugins.
How did you get the green screen scenes so clear and unnoticeable?
CL: The green screens to be honest were really bad. The conditions which they were shot under were just horrible, it was night, it was raining, there was mist and fog all over the place and you’re shooting through glass…So I would have to say we just had really good compositors.
What tools did you use for motion tracking to make it all look seamless?
CL: Our motion tracking, or camera match moving was done in Boujou and Matchmover.
I loved the frozen pond piece it was looked so simple but so effective, how did you do it?
CL: It wasn’t simple, (he laughs.) All of that work was done through Houdini which worked really well. We had a fantastic Houdini crew working on that, I can’t say enough about them. Defining the look of it was so difficult, it’s hard for people to articulate what that would look like with words. But we locked it down. It’s in the trailer and everyone seems to like it.
Did you do the entire bridge sequence or just the moving of the bridge, if it was the entire sequence, did you have trouble taking the scene day to night?
CL: No we did right the way up until when the bridge slams down on Alcatraz… at that point we handed it over to Weta.
How long were you working on the film for?
CL: Me, wow, we geared up in April/May 2005. So a year and some change.
You guys must have run in to some problems at some stage, how did you overcome them?
CL: The creative side was the hardest with all the changes in directors. The only way to overcome that is to be flexible, ultimately at the end of the day they are our clients and we give them what they want. It’s tricky because when someone asks for the world, you have to deliver it. To make a large change, there aren’t enough processors or bodies to render it…so that was a frustration for us. On the technical side…getting something as large as that bridge rendered…was a miracle, that and people making changes, but hey, we managed to do it. We pulled it off. We were just prepared to do anything and everything, but sometimes you have to be prepared to say no, which is the most difficult thing.
How do you feel when you watch the film now?
CL: I can’t watch the film yet, it’s too fresh for me. You have to re-live your whole year/year and a half. You lose the illusion of film. You stop enjoying it. You tend to find that the films you enjoy the most are non visual effects films.
How did it feel to finish the film?
CL: Relief, empty, I think everyone goes through that. You work so hard up until the last minute and then all of a sudden there’s nothing left and it’s done. You have a great feeling of loss.
Are you happy with it?
CL: It’s too soon for me to say. I need to look at the show again in a year. There are so many shots that we did that didn’t make it in for whatever reason. I’ll let you know in a year.
Is there anything that you weren't happy with?
CL: We didn’t have a ringer in this one no, there usually is one or two that you look at and don’t feel they should be on the screen, but not in this movie, so I would have to say no.
VFXTALK: Did you have to develop/ customise anything to get the job done?
CL: We had to change our pipeline a lot, just getting that amount of data and it rendered… the geometry, the millions and millions of polygons that the bridge was…it’s goofy huge.
How many shots did you do in total?
CL: Around 175, there were a lot of omits. We would put beautiful shots together and all of a sudden they would be gone from the edit, which is so frustrating.
How long did it take to render?
CL: Beauty renders of the bridge, anywhere between 2 and 12 hours per frame. The shots were long, they were between 8 and 10 seconds per shot…so the camera holds on it. Rendering was just a bear.
How real/close do you actually feel you got to what they visualised?
CL: It’s funny, because like I said the vis was changing, which takes you in a new direction, but once we locked on it the shots never changed so I felt better about that. We originally were supposed to be finished in October last year…we didn’t get the plates turned over to us till January 06. The good thing about having them so late, which was the only good thing I might add, was that we got to revisit the action of some of the shots and see how it worked. Once we had it locked down nothing changed…so hey, something must have worked!
Is there anything else you would like to add?
CL: London is in a really good spot right now, especially Framestore, it’s an exciting place to be. All of the major studios are now looking to London for their quality work, which is great.
Do you have any wise words for our community?
CL: I remember being there, trying to get in to this industry and it was really hard. The schools didn’t exist back then. The only thing I can tell you guys is that you have got to be persistant, stay on top of it. You have to keep on submitting your work, send a new reel, send a new reel, send a new reel.. Go to the talks, find out who’s giving the talks, go up to them…you can’t be shy about these things, introduce yourself, talk to them, get in to their radar and stay in that radar. Persistence, that’s the only thing you can do!
On behalf of all of VFXTalk, thank you ever so much Craig.
CL: Not at all, it was great and good luck to you guys out there.
Written by VFXTalk