Be part of the VFX neighborhood by changing into a b&a Patreon…and get bonus content material!
How ‘Emancipation’ VFX supervisor Rob Legato designed the Will Smith vs alligator scene.
At one level in Antoine Fuqua’s Emancipation, the character Peter (performed by Will Smith) seemingly escapes his enslavers, solely to confront an alligator in a swamp.
Visible results supervisor Rob Legato labored with a number of departments and with Legacy Results and Wētā FX to deliver the alligator assault to life. This included capturing components in an underwater tank with an LED display screen set-up to venture underwater Unreal Engine-made swamp footage.
Legato shares his ideas with befores & afters on designing the visible results for the scene and making certain it was a stunning and terrifying second within the movie.
b&a: What have been a few of the first conversations you had about that scene with the filmmakers?
Rob Legato: Simply earlier than I met with Antoine, I downloaded some footage of a crocodile snatching a leopard from the shore. It’s a type of scenes the place you’re taking a look at one thing pretty harmless, after which in a heartbeat, it turns into this super-violent factor the place the leopard is ripped from the land and into the water. That was my impression about what I needed this scene to be like, and that’s what Antoine appreciated as properly. That was the thought, that we have been going to do one thing that was surprising and tremendous quick.
I imply, it’s somewhat bit anticipated as a result of he’s within the water the place there’s alligators. We determined to movie it in such a method that he was about to be in complete freedom. He was hiding out from the trackers, you assume that possibly he’s free and clear, after which he will get snatched.
b&a: In these preliminary conversations, did you additionally begin speaking about how you’ll carry out the precise stunt and pull off the shot?
Rob Legato: Nicely, my strategy was to do it as a combination of every little thing to create probably the most fascinating model of one thing. You can simply say, ‘Okay, it’s going to be a CG alligator, CG Will.’ Sure, you may try this. However you additionally need a few of what the actor brings to the get together. You need a few of the haphazardness of dwell motion images that’s not so excellent.
The plan was that we’d shoot him in close-up in a swamp wrestling with one thing. That method, you get little bits and items of issues that, for the 18 to 24 to 36 frames they’re on, persuade you that it’s actual. But when a shot was 48 frames, maybe it will most likely not as convincing. Or, 5 seconds, it most likely wouldn’t, both.
I like the thought of doing dwell motion, actual place, underwater tank, LED display screen, animatronic and CG, after which glueing all of them collectively. You possibly can create an editorial movement to cherry-pick the very best moments. CG could be very tough, in that you are able to do something, however something tends to look somewhat too animated, studied, excellent. So that you attempt to sandwich it in when you shoot the dwell motion.
b&a: Is it a sequence that you simply needed to have boarded and even previs’d?
Rob Legato: No, not boarded. I’m not a storyboard fan. Generally meaning you might be pressured to attract one thing, and that drawing that you simply’re pressured to create is just too stiff. It might probably power you into one thing that I believe makes scenes like that too stiff, too regimented, too deliberate, the place what you really need are comfortable accidents. You need the shock.
It’s one thing I realized from Lion King and Jungle Guide, particularly in coping with animals, is that you simply undergo YouTube and every little thing else and also you cherry-pick. That piece of motion seems actually fascinating, this one seems actually scary. You begin to categorize them. Okay, that may be for this second, that may be for that second. You litter the scenes with actual habits that you simply then glue collectively. Weta Digital, who have been fantastic to work with, by the best way, created these animations that glued collectively habits and motion that we appreciated.
By way of previs, I like to try this actually right here in my basement. I previs it and stroll round with a handheld digicam after which say, ‘Nicely, this shot will solely work for this second. This shot will solely work for that second.’ You break it up as you’ll have when you have been to {photograph} it for actual. You get this angle and then you definately get that angle and also you get that angle. You simply carry on capturing it till you decide every little thing aside, versus attempting to get every little thing in a single. Generally getting every little thing in a single is the best way to go, after all. However most motion scenes are a collection of cuts.
b&a: How was the dwell motion filmed?
Rob Legato: We shot just a few pictures in the true swamp with a Will stunt double arising out of the water and stabbing at a Legacy Results-made tail that was swooshing backwards and forwards. That turns into the above water second.
After which we matched regardless of the stunt double was doing by capturing Will in a tank. Curiously, we constructed an aquarium in entrance of a TV set, which was mainly an LED wall with the swamp background and clear water for Will to have the ability to swim round in comfortably, but nonetheless be suspended underwater. After which we had the comfort of having the ability to shoot exterior the tank to do our digicam work.
Then we had a ‘sandwich job’ to do, the place we’d put some CG ferns within the foreground. So we had Will within the water, after which the background behind him, which was an Unreal Engine model of the swamp. You see simply sufficient to fulfill the truth that you’re seeing one thing. You don’t overly spotlight it, you don’t overly gentle it, you don’t do any of that stuff. The truth is, you need the murkiness and also you need all that stuff to confuse the attention.
Within the tank, we shot with an animatronic from Legacy. Then we shot some stuff the place the dwell animatronic was going to get replaced, after which we shot some with out it. Then extra with the stunt double who might bounce in and dive and spin round. Then, the CG alligator was added to that. For some pictures, Weta did a full CG Will and full CG alligator, though that was used very sparingly. There’s solely a handful of pictures that I used it in. They have been all the time in the midst of motion.
b&a: That LED display screen strategy sounds actually fascinating.
Rob Legato: It was enjoyable to do. I don’t assume anyone’s ever completed it earlier than. I wrote to James Cameron and stated, ‘Hey, you would possibly need to do that together with your underwater tank.’ It was 25 ft lengthy. A transparent entrance, clear again, so you may put an LED display screen in there. You can act such as you’re swimming after which simply change the background to make it appear like you’re swimming and touring 50 ft, when you needed to.
I assumed, ‘Nicely, that is the best way to do Aquaman and all these different movies the place you’re suspended such as you could be underwater.’ You can then swim what seems to be nice distances by the phantasm of the background transferring and nothing extra, since you might swim in place.
It looks as if it’s a cool method of making these scenes and utilizing an LED wall to get in-camera work. It actually helps as a result of there may be some murkiness within the water, there are bubbles, there’s all of the micro issues that let you know it’s actual, as a result of they’re. After which to get it in-camera and never need to attempt to matte it.
b&a: I believe that’s your modus operandi in lots of movies that you simply’ve labored on, as you’ve talked about, the place you do it so many various ways in which, actually, the viewers is a bit uncertain of the way it was completed.
Rob Legato: Sure, actually animation-wise, that’s why I all the time need to shoot a dwell particular person doing one thing. I don’t desire a dwell particular person on wires doing it, as a result of that’s additionally synthetic and also you are likely to see that as properly. So I attempt to do as a lot in-camera as doable with the true physics after which minimize my method out of it, versus attempt to get one clean, physics-defying swing or bounce or one thing that now turns you into form of a superhero when you see it multi functional than a typical motion scene.
Be part of the VFX neighborhood by changing into a b&a Patreon…and get bonus content material!
Associated
