You were saying off-camera that repeating patterns present unique challenges for effects.

Yes.

Tell me about that.

Even on this set, we had, when you look at it initially, you kind of go, oh, it's just a big set.

But then you realize it's almost a mathematical problem.

We have the bridge repeats every 120 frames and degrees.

Then you go and you look at the pipes, and the pipes repeat only every 180 degrees because you wanted to, it needs to play, it's part of what the story is.

Then you start realizing that when you're seeing repeating patterns, you look down the silo, it becomes very quickly, it just reads like a very long pipe with a spoke kind of And you need to break it.

You need to try and realize what I can do to still retain that kind of scope of a city and look like the way you look at Fifth Avenue and kind of go, I can recognize a city.

What kind of triggers that in the horizon all the way up to close to where you're looking from.

And it's just about the movement of the crowd, how to break the textures of the concrete, the light, making it feel like it's not, you know, we couldn't have neon lights everywhere and just put distance signs, so we had to play with our own lighting and just break all those patterns with textures.

For example, our farms, as you go further back, we let them kind of bloom and kind of pollute the atmosphere of the silo and just kind of slightly feel like, you know, a distant grow light is kind of a bit stronger and different colors and bringing magentas and cyans into the horizon, but not have them in the warm, closed environment of the silo.

I've been watching, walking through the sets here and they're so immersive, so all enveloping and yet you were showing me the show reel and you're still having to provide many times 100% of what's in the frame.

Yes, we were obviously limited with how many variations of this world we can build.

And when you start looking into it, we needed to create different, people need to constantly feel like they're moving up and down this place and feel different, like you want to be able to almost allow the audience to localize themselves in the space.

So, a lot of times when we moved on the floor, we had to start bringing in our blue screen.

As you can see, we have a lot of it, but just to retain the scope so we can have someone continuously walk from a floor and land on another floor without repeating the same set, without feeling that, oh, it's just one set repeating itself.

We wanted to constantly make sure that you feel like, oh, it's a new location or I'm familiar with that from the previous episode.

I'm now, I know exactly where I am, which is kind of an interesting play with how you make people that don't know the silo slowly get familiar with it and almost feel like by the end of the season, they could read, they can very quickly go, oh, that's upstairs.

And you must end up with the same thing, like the silo must actually exist in your head.

We became the map keepers on the silo because, you know, we look at the blue and go, oh, if he looks this way, it's the wrong bridge.

You need to rotate 120 degrees.

The eye line from this bridge, when he looks 10 floors up, actually lands on the wrong bridge.

So how do we turn him around to make sure he's looking at the right person and he's looking back?

Because we had a lot of those contacts of continuity between our set and what's happening above us or below.

Now, I think a lot of people think of special effects as kind of just a toolbox full of tools.

But I know that every shoot brings its own unique, specific challenges.

What has been some of the biggest challenges in bringing the silo to the screen?

The main, I would say there's been quite a few that kind of required their own solutions.

One of the things about this silo is that there's no straight lines, which doesn't seem to strike you originally.

But when you start dealing with crowd movements, it becomes a major thing you need to solve because suddenly you go, oh, someone walks down a stairs, a set of stairs.

Spiral staircase introduces a completely different body language of what it is walking downstairs.

Oh, especially if they're on the inside or the outside. So we mapped it kind of like we had rules for the silo.

But when we break the rules, someone's running up, he doesn't care about the rules of keep right or keep left.

So the stride of the stairs became a major thing we had to solve how to capture.

And we had to build our own set of stairs just for the motion captures to be able to have the correct physicality of if it takes two steps down each one of the steps or the very narrow inner kind of stairs.

And it creates a completely different angle for the shoulder, the way your eye line looks when you kind of you're careful not to stumble down.

And we had to build a whole texture of that.

Then you have how do people step off the bridge or go on that small platform between the two different flights of stairs.

All those movements had to be captured so that we are not limited by our crowd and that we can continue continually see people walking on the outer rotunda.

So every curve suddenly changed the physicality of how we can capture our movement.

And that's crazy.

It hadn't occurred to me.

So did you end up like parceling out?

Because the stairs are getting wider, of course.

Did you like separate it into zones?

So, you know, I'm walking zone one, zone two, zone three.

We had to.

So on this season, we kind of we were taking it to the next level and we actually built our stairs in multiple ways so that we can go onto the bridge, off the bridge.

And we had a proper flight of stairs because we went to the bigger motion capture set.

Last season, because of limitations of how you can create the volume of motion capture, we created flight of stairs that are the different strides.

So we just had the very narrow inner stride, the middle one and the other one and had our stunt team kind of run up and down, walk.

They had to.

You realize that very quickly, some things feel forced and don't quite work.

And so we had to rebuild our motion capture approach for the silo.

And it's how people go around each other, even when you walk in a curved line.

So you need to push around.

And that's what the main thing for the silo was to not have what I call the classic stadium crowd, which is we have waving, we have cheering and flags.

You wanted to not have the train station effect of people just walking back and forth.

So we have a lot of stories.

If you look in the back of our silo and in some of the episodes, you can see a bunch of kids playing hide and seek because we wanted to feel like a city.

You know, it's a place people live in.

So it doesn't always have that movement.

You have rush hour, early morning, high noon.

Every time they had a different approach.

And then we had bespoke events.

We had a race.

We had a festival.

People need to react.

Someone is telling them now we're singing.

So all those things layered up on our crowd build.

It became a major challenge for both seasons.

And in season two, you've got some underwater stuff.

I'm curious what challenges are presented by effects underwater.

The idea of closing the door and turning on a garden hose got rejected.

So we had to build a way of approaching the silo in a way that it's flooded.

It's a new design in a way.

And everything you think about, a curved world suddenly becomes quite different.

The water has its own lensing effect.

So all of our round lines become...

You're accommodating for the angle of refraction in water.

Which is one of the things when someone says, oh, we do underwater, that doesn't take care of it.

And then you start thinking about it and you start testing and you go, oh, that is something we need to take into account for our camera tracking, for interaction of Juliet and how she deals with it.

So all those refractive elements are things that we need to deal with.

We need to figure out, give it a look that is dirty, but find ways of...

Our silo is very dark as well.

So we need to play with what tells the story of a flashlight rolling on windows.

So we cleaned in our model, we created clean spots on windows so that you can catch your own reflection.

Just to make it a bit scarier and a bit more like a set.

So I'm curious, given that you guys had to shut down for the strike, if during the time the strike was happening, did that give you time and perspective to think about what you were going to do when you came back to it?

Did it give you a kind of different...

Yes, it's a problem when you sit at home and you think about what I would have done.

Actually, we should do it differently.

Who do you talk to about it? You sit at home and go, that's a really good idea.

I should remember that and put it in your small notes.

Potentially come back to it.

We kept working on some elements.

So we were lucky enough to have finished shooting part of season two.

So that really helped us.

So you were able to actually continue.

We could continue and that actually gave us a lot of ideas on what can or should happen.

There wasn't such a time crunch because there's not more shots arriving.

Exactly.

We had a section of it.

It was a bit of a challenge because we haven't finished shooting.

Usually you want to see everything, but it allowed us to really wrap our heads around some of the challenges we're going to deal with.

With the underwater silo, with the underwater steps, which we had to build.

And kind of small things.

Ropes underwater compared to air hose.

Oh, right.

So, you know, when someone goes in, a rope is heavy and sinks, the air hose floats.

And then you need to extend it.

We went to shoot in the Pinewood tank.

Oh, you have to give your digital cables their own physics.

All physics and all movement.

And then you frame and you go, oh, that rope is pulling this way.

The air hose is pulling that way.

Let's wrap our heads around what she's interacting with.

Oh my gosh.

A lot of it is extensions because we're going so deep underwater that really we couldn't put it into a normal water tank.

A lot of it had to be cheap.

We had to put her sideways.

We had to use water movers to create movement when she's static.

So we had divers with...

Fans.

Fido's.

Pushing water. Very effective because it pushes the bubble.

It pushes the air.

And then you feel, oh, I'm reading movement.

Then we have to create our parts to help that movement by having the backward move with it.

Well, so after all of this time, you guys are wrapping photography in just a few weeks.

But your job doesn't end there.

You've got a lot more stuff to do.

Until you get to see it at home.

And maybe a little bit after.

And you guys are working on effects almost right up until the airing.

Is that true?

You push.

There's a lot of effects.

And even on the smaller things when it's just, oh, there's something in the window.

There's a small fix in the background.

A lot of those things, we just work on it until you manage to deliver the entire season.

I'm curious.

Is there a shot that you feel a tremendous amount of pride over?

Or like that was one of your favorites of the season?

On season one.

Season one or season two.

Season two, it will come.

I'll be proud of it when it's done.

We sign off the last pixel. But yeah, I think I'm very happy with the generator room as a sequence.

We have a few shots there that are for me just.

That sequence is one of my favorite engineering sequences in a thing.

I was saying to you off camera, a lot of time engineering sequences are strange belief is a nice way to put it.

And that one felt really grounded in the world.

Rebecca's performance is amazing.

What you guys gave her to work on feels so real.

And yeah, it was very important to give her something to actually interact with.

It's very easy to drift into, oh, let's just put everything on blue because we're already doing so much blue.

But it was really about how do we connect?

Let her have something to work with.

Let all the cast be in the same place.

And at the same time, because of the way the set was structured in height, it was actually split into three different sets.

So a lot of shots you had to solve.

How do we get someone to step out in one set and connect with Juliet sitting at the top of a massive...

In a totally different room.

Which is a different set, which had to be done because of safety and height.

Yeah.

We hit the ceiling at some point.

But how to connect it and how to not limit the director that wanted to be able to move handheld, the energy of the sequence required as motion control was ruled out because of placement weight and how to do it.

So it became its own kind of challenge of how can we drive the sequence and make it feel very real. By just vibrations and modeling, there was a lot of animation work that went into making the space, even though it's a background.

Just you will see there's a lot of vibrations.

We looked at a research paper that shoots machines and allows by motion analysis to see micro vibrations that are actually not visible.