CG spiders, webs and even a crashing car formed part of the studio’s work on the film.
On director S. J. Clarkson’s Madame Web, visual effects studio beloFX was called upon to craft a number of signature VFX moments.
These included moments with the film’s defining spider, views of spiderwebs, the spidery mask of the Aranas, and the crashing of a car off a bridge.
For each of its VFX moments, beloFX had several design challenges to overcome, while also working with production to solve crucial story points.
Here, beloFX visual effects supervisor Oli Atherton–who worked with production visual effects supervisor Mike Brazelton (also hailing from beloFX)–breaks down those key VFX shots and how the studio achieved them.
b&a: We see beloFX’s work right from the beginning of the film where the mother is in the Amazon researching spiders. There’s the web there you had to do. Tell me about those shots where you’re seeing her right in front of the web.
Oli Atherton: That was a very interesting shot because it came into the edit and went away from the edit in a few different forms. At the very beginning, this was going to be a big shot with a descending spider and the camera moving around. There was a much larger segment, another 2000 frames at the head of the shot, that we tracked and blocked out a spider for. I think, for the better, it was cut down because it’s a fairly long shot and they didn’t want to reveal the spider just yet, since she’s still in the process of trying to find it. The goal was a unique and recognizable web so that the character sees it and thinks, ‘Oh, this is the one. I think the spider’s been here.’
We built a procedural web that we could string on trees and fences and so on. But ultimately the director S.J. envisioned the web as resembling a cracked windshield. She sent us a reference image, and we followed its outlines with splines and then modelled it, and that became the center of our web. Our procedural web was used on the outer edges to string it up, while the core was designed to precisely match this shape, creating this distinct and recognizable motif.
Then there was a significant effort into getting the lighting and compositing right to make it feel natural. We explored a lot of spiderweb photography references. There were also a lot of spiderwebs in the garden, and I was just going around taking pictures of them in different lighting conditions, holding things up to block the shade and spraying water on them. Mike Brazelton and I both agreed that webs are very delicate and beautiful, yet at the same time, strong – and so we had to capture that. We found they always looked best in photography when they were lit a certain way and so we started to recreate that in some test renders as we dialed in a look that Mike was happy with.
b&a: It’s funny, isn’t it, because I’ve covered so many Spider-Man films and I feel like I’ve never really asked about the art of doing a web and what the very intricate details are that are required. Things like highlights and just the fact that it’s so thin, that can be a challenge, I imagine.
Oli Atherton: It was. Depending on what lens you’re on, how close you are, and the lighting, it’s easy for parts to completely disappear or the opposite. Then if it’s forced too much it can start to look fake, very fake. We did go through quite a few iterations with Mike and the director as we were all working to find this balance.
b&a: What about when you came to do spider shots?
Oli Atherton: I was gunning hard for the spider really early on. It was something that beloFX wasn’t going to work on initially. We had to first demonstrate to S.J. that, hey, we’re going to make this look fantastic!
After our first base asset was done and rigged, I looked through some online clips of people holding black widows – that was the species that most closely resembled the spider we were creating. I sat with our animator and said, ‘What things make this feel like a spider?’ Ultimately we found that it was the clumsiness, the way they fight against gravity. When a spider is walking on a hand, bits of it slip, small movements like that. We did a bunch of tests capturing these types of movements, and eventually built that out into a long sequence.
The director wasn’t keen on traditional turntables. So instead, we crafted a sort of animation hybrid – beginning like a turntable but evolving into the spider coming alive, dropping down a web and landing on a plinth. I suppose it was a cheeky way of getting a turntable in there because ultimately that’s an important part of getting an asset signed off. But we were able to include those animation elements quite early on and get a feel for how the spider will move. This approach was well-received.
The asset itself evolved considerably, specifically the amount of visible hair. They were trying to strike this balance between feeling realistic, but also scary, while also avoiding a look that was too cute or friendly. So it was all about trying to get that balance with the hair, the placement of the eyes, the size of the eyes, how much glint you see in the eyes, the tones of red, and so on.
They had a small maquette, but, of course, it translates quite differently. It was a somewhat metallic maquette, and its appearance varied a lot under different lighting conditions.
Then there was the issue of colour. Should it be Spider-Man red? Should it be the red of Madame Web? Eventually, they landed on Cassie’s jacket colour, a hue she wears throughout the film. The consensus was, ‘It should look like this.’
b&a: I actually think it’s an interesting conversation just in terms of what a visual effects studio typically has to do to get to the vision.
Oli Atherton: Yes, interestingly, one of the early animation pitches involved replicating the spider from the first Toby Maguire film. We copied that animation exactly, because I thought it’d be a really nice homage to include in our work.
b&a: For the rainforest people, the Aranas, what kind of mix of make-up effects and final visual effects work was required there?
Oli Atherton: During the shoot, they wore a costume that was essentially body paint, more like a body suit, complemented by face makeup. Over that, they had this mesh made of fine twisted vines. They built their dresswear to blend into the rainforest. However, one thing that happened was, the facial makeup just kept coming off due to sweat. We had quite a few shots where we needed to put makeup back on them, so we scanned a few actors, lookdev’d the look and rigged it all, especially for Ezekiel in the cave.
In the end, that approach was scaled back. However, we ended up working on the headpiece. Although they had a headpiece on set, they decided to shoot without it. They sent us images, and it resembled more of a sack or a stiff fishing net over the head. Early on we collaborated with our concept artist to explore a few more variations. In that process, we looked at the original Spider-Man costumes and leaned into those.
We essentially tried to subtly pay homage to the Spider-Man costume. When we sent it over, Mike loved it and encouraged us to do more, which got everyone quite excited. So we dove back in and really leaned into the very early Spider-Man costumes in the comics, aiming to recreate those shapes. Then we looked at the movies as well for additional inspiration.
Then S.J. expressed interest in revisiting the very first iterations of the Spider-Man costume. That was fascinating because this idea evolved creatively throughout the process. It was pretty exciting to, in a way, be part of bringing the very first Spider-Man costume to cinema.
b&a: Was it a difficult thing to realize as a mesh on top of a face?
Oli Atherton: We certainly had to meticulously body track it. Then the tricky bit came when we needed to translate these 2D concept paintings into 3D space. It could look great, and match the concept from one angle, but as soon as he turns the silhouette breaks. Another thing was working out how tight the mesh had to be. Did it need to be as tight as Spider-Man’s mask? One note we received was that it needed to more closely resemble the comics.
b&a: When Cassie in the car goes over the bridge, what did they shoot for that scene?
Oli Atherton: They shot this in Boston, standing in for New York. Almost everything you see in the background is DMP. We did use Unreal Engine here, taking some city map data and on set photography that Mike’s team had done of the New York skyline. S.J. wanted certain things like the Empire State Building to read better, so it’s a bit more noticeable.
The car was fully CG, but they did film a car going over the bridge. Initially, they had a car on a rig and pushed it over the bridge, but it didn’t end up being dramatic enough. To enhance the scene, we scanned the car, modeled it, and tried to match the damage as closely as we could. We then aligned it with the practical car for lighting, and experimented with a few different animation timings. The water also needed to be CG for the realistic interaction with the car.
Then there were the railings, which also needed to interact with the car. So we removed the railings, recreated them in CG and animated them to buckle a little. There was a bit of sculpting and fiddly animation to get that to work, and then some dynamics just to carry it through.
In the finished shot, most of what is visible beyond the man in the foreground is CG, including a silhouette of Cassie in the car. At first, we weren’t meant to see her there, but now that it was CG, we could easily add her in. This also meant that we could amp up the impact of the car hitting the water. Initially, they filmed with the water much closer to the camera. Now we could push the water away to have it play longer and be more dramatic.
b&a: Again, I just love that it’s VFX problem solving.
Oli Atherton: Yeah, especially when they wanted to get Cassie in there. They did have a scan of her, but it was only a proxy one. That prompted a conversation about whether we needed to lookdev a fully digital Cassie, or if we could keep her in silhouette. Sometimes, when you’re working with time and budget constraints, these types of practical compromises are necessary and can open the door to creative and inventive solutions. I think when visual effects works like this, and everyone gets on the same page to solve a problem, it’s great.
