Think of the technological advancements we’ve seen in moviemaking in the 125 years it’s existed as an art form. From the kinetograph, patented in 1891, to talkies, colour and special effects throughout the 20th century, the story of movies has always been one of technological one-upmanship.
Like the digital technology that now drives movies, the 21st century has seen even more milestones, from realistic colour 3D to 4K, cloud computing data delivery and beyond.
But among the changes technology has wrought, the most pervasive must surely be computer-generated visual effects – the backgrounds, foregrounds, characters, action scenes and physically impossible images created inside the computers of VFX houses the world over, the director and actors on set working with a few simple interiors and props.
Today almost every huge, midyear action adventure movie (the business model every Hollywood studio is vying to conquer) is more cartoon than movie, a handful of live action clips surrounded by thousands of effects shots and elements.
So if we credit “The Horse in Motion” (1878) with the birth of the moving image, “The Jazz Singer” (1927) with the introduction of sound and “Avatar” (2009) with digital-age 3D, where does the credit lie for the modern methodology combining actors and green screens with completely animated backgrounds?
By the early 2000s it had been done here and there in key sequences, set pieces or characters, but the first film to do it entirely didn’t light the box office on fire, make an A-lister out of its director or garner critical acclaim.
An action adventure thriller set in an alt-history wartime America that combined the swashbuckling aesthetic of matinee serials, the romantic, soft-focus sheen of Hollywood’s golden age and a steampunk sensibility to machinery and action, 2004’s “Sky Captain and the World of Tomorrow” was an unwitting manifesto to the rest of the industry; this is what VFX can (and will) do.
While stars Jude Law, Gwyneth Paltrow, Giovanni Ribisi and Angelina Jolie filmed principal photography with writer/director Kerry Conran using only costumes and a few props and half-built interiors at Elstree studios in the UK, the purpose-built animation facility in LA’s Van Nuys neighbourhood led by his collaborators Steve Yamamoto and Steve Lawes built everything else, one pixel at a time.
Moviehole talked exclusively to Conran on the 15th anniversary of the film’s release to ask how it went down, where he went afterwards and the true influence of his handmade classic.
Talk about how the movie came together.
I went to Cal Arts [where most of the early brains trust behind Pixar studied] which is known more for its animation than anything else. I had always loved animation, certainly the way it was made – you’d create a background and then use cels to animate the characters on top of that.
Around that time accessible compositing software like AfterEffects was just coming out, so I wondered if there was a way we could apply the techniques of animation to live action. The first couple of tests I did it was really simple stuff, just taking old photographs and putting people into them. it sounds trivial, but it became more ambitious where you can dimensionalise the photos and if you planned it out it could look compelling and convincingly real.
I was experimenting with only black and white vintage photos, and it made it easier to combine modern day live action, the compositing just wasn’t as complicated.
From that I started working on a short film, really to test out the concepts. I was using Electric Image at the time and my machine was so low powered I’d have to render out the robots one piece at a time, literally an arm first and then a leg, because the computer couldn’t handle it otherwise. If I had 100 robots I had to do each one piece by piece and reassemble them after they were rendered.
So it was difficult but the fun part was seeing these beautiful vintage photographs that had a production value, a hue, a scope and a scale I couldn’t possibly achieve. That’s when I thought ‘this could work’.
That’s what I set out to do, I wasn’t trying to reinvent the wheel, it was literally just a means to an end, a way to get these ideas on screen with the tools that were becoming available.
That’s how it began, but it changed quite dramatically once we got funded. I’d originally looked to spend about $3m total and the film would’ve been completely black and white, done with only photographs. Almost instantly that went out the window and even though we still used a lot of photos we moved to a CG virtual environment.
In order for it to work we had to know every background in advance. We ended up having to previz the entire film. Previz was certainly used back then, but we did it for literally every single shot because we had to.
Nowadays big productions with a lot of VFX will build animatics of sequences to plan them out, but you had to do that for the whole movie?
Completely, and it was even a little bit more than that, because today’s previz gets completely thrown away and then the effects shops rebuild the sequences from scratch with higher resolution models. We kept everything, we worked on top of the previz for the whole movie. We might have changed some of the models with the higher res versions but a lot of the animation at least was blocked in.
You didn’t have the luxury like you do today were you can see [the virtual set], you can literally put it to your viewfinder and compose a complete shot.
So you consider that you were actually drawing on the past – being animation – not inventing the future?
Utterly so, my influence was probably more Disney films and German expressionism and things like that. So it took older styles and translated them with very modern tools and they were up to the task, it did what we needed it to do.
Considering it was all CG with just a few props and costumes, what actually cost US$70m (the published budget)?
One of the big misnomers about this film is that it cost seventy million dollars. I don’t even know where that came from, I think it was used as a ploy to get more for the sale of the film.
The original incarnation was supposed to cost $3m million, but when we had to go to colour and virtual our budget went to about twelve million, then increased to about eighteen million dollars.
The film was ultimately sold to Paramount, and they were led to believe we could finish it in May of 2004, which was never a possibility. So when they realised it the bond company threw in money and that’s when all of these other effects companies got involved.
If there’s anything I wish were better known about the making it’s that we had three veterans but the rest were really new to the industry, some just out of high school. [Visual effects artists and Conran’s original collaborators] Steve Yamamoto and Steve Lawes weren’t paid a penny for their work early on but they felt invested, they felt like they were doing something different and new.
[When Paramount signed on, the bond company insisted sequences be sent to external VFX vendors to meet the release date – Conran and his team had worked on about 70 percent of their shots, which included 25 minutes of finished movie.]
Was there a sense on set of how it would work?
From our first day actually, the camera crew in London were incredibly polite but privately very skeptical. We were shooting these scenes against a blue screen and there’s nothing else there and they just thought ‘there’s no way this is going to work, it’s going to be a disaster’.
That first night we sent the footage home to Van Nuys and Steve Yamamoto really quickly put together a basic composite of the previz with the footage we shot and shipped it back to us for the next morning.
So I got to show that to the camera crew on the second day before we even started, and it completely transformed their opinion of the project. It was obviously unfinished and crude, but it was amazing. You saw these actors actually interacting with the set and these virtual backgrounds and I think it was kind of a revelation [for the crew]. That was probably the moment we thought ‘This is going to work’.
It was one of the first times a movie had been sent out to so many external VFX vendors to complete shots and sequences. How hard was that to wrangle?
It was difficult. It really fell on the three veterans [Scott Anderson, Brooke Breton and Joyce Cox, VFX supervisors/producers assigned to the production in order to facilitate the different vendors]. By the time the other VFX companies came in we were fairly late in the whole process and we’d always intended to do the whole film ourselves.
Our approach was pretty naïve, and a lot of what we did was unconventional. We did whatever we had to to get shots done, so the VFX companies weren’t quite prepared for the workflow we came up with. It was of a volume that kind of stunned them initially, they didn’t think it was possible.
So we walked them through it . We knew it’d be difficult for them to match everything so we did it for them. Steve Lawes provided them effectively a few finished frames from the shot and the software presets he was using. It was basically sort of a shot kit – whatever sequence they were working on, we did as much as we could to explain it and provide them with the settings.
Some companies took to it really well. ILM was quite amazing and gracious and I think they appreciated a lot of things we were doing and even adopted some of them later, because we were spending a lot of money and effort building things out that they didn’t have to.
[Adelaide, Australia-based] Rising Sun also really took to it and was great to work with. Some companies didn’t succeed and we had to take the stuff back or put a guy with them to help them through the whole process. Some struggled, I think just because it was too ambitious or you had to really embrace our methodology or you would die.
Even though big effects movies nowadays subcontract so much of the VFX out quite readily, was it hard to have that imposed on you?
Yeah, that was tough. The way we were doing it, it was a very handmade film and we had a strict schedule and all that, but we had what we felt was enough time to finish it.
But it became a rush kind of thing, which is more traditional. It’s just that when Marvel has to finish a film they’ve got the resources and money to throw everything at it so the quality is obviously the highest you can get.
We weren’t as lucky, a lot of the stuff we were getting back or doing ourselves was so rushed towards the end it was difficult to maintain the style and quality. Some stuff is painful to look at because it’s so crude by comparison – crude even in our methodology back then.
I still really strongly believe in using previz, and what guarantees you a good result [when using external providers] is making sure your production design’s in place so you can trust these companies to pull it off and probably give you back something better than you imagined.
They have these amazing artists that work at all these places that are probably better filmmakers than the people making the films. It’s a totally different world.
Would you be more open to making it using external providers today if it was more your decision to do so?
Being hands on with film is the fun part of it. The stuff I’ve been working on recently is similar in that way – it’s different in terms of the technology but to some extent it’s just the way my mind works. I am a little bit more hands on and tend to keep everything close to the vest.
But working with outside companies wouldn’t bother me anymore. I think with Sky Captain I exhausted myself, so I’m cured of having to have my hands on every aspect.
It seems that as a blockbuster director nowadays, you shoot actors on sound stages to get the principal photography in the can and after that you’re not actually making anything, you’re just sitting around waiting to approve stuff other people have made.
These days particularly. Steve Yamamoto has done most of the previz work for Michael Bay since Sky Captain, and guys who are doing that work are more or less co-directing those films.
When you see these big, complicated action sequences, directors are approving them but other people are sort of envisioning them. Obviously, the director plays as big or small a role as they want, but when you know what VFX artists can do you don’t have a director sitting there dictating every little move or shot or sequence. They’re made by really talented individuals and then the director tweaks it or ultimately approves it.
It’s changed the job description kind of fundamentally.
Entirely, and I think that should be credited. People don’t really fully appreciate what effects companies and effect artists do because they’re filmmakers.
Nowadays any actor who’s been in a big movie knows all about performing against a guy holding a long stick or nothing at all in some cases, but back then it was very new. How much did you have to hold the hands of stars Jude Law, Gwyneth Paltrow, Giovanni Ribisi and Angelina Jolie?
Almost not at all, they were incredible. For them the nearest comparison they made was that it was more like theatre, because on stage you’re not used to having everything there and you’re meant to act in an environment that’s not represented.
So it did call on them to be a little bit more aware but the things that really helped was all the preparation. Before every shot we watched the previz so they knew exactly what was happening in a scene.
[The camera crew and actors] had to frame for things that weren’t there and we had to make sure there was a one-to-one translation from the physical set to the CG set. So between that and just blocking the scenes out, I don’t think we had any issues with it. Gwyneth would literally just ask ‘Where do I start? Where do I stop?’.
Jude was so engaged and intrigued by it all [Law came on board early because of his interest in the concept and was pivotal in lining up the rest of the cast and subsequent studio backing) he didn’t have any preconceptions so it didn’t faze him. He fell into it, it just felt intuitive and comfortable for him. I know Giovanni was so taken by it that he actually enrolled in Maya classes.
Angelina had just come off the second Tomb Raider film and I only had her for two days, so she didn’t have the benefit of sorting through the whole process like Jude and Gwyneth did, so she was immediately thrown into it.
In one scene she was more comfortable entering the scene from a different direction than what we’d set up, and it was difficult to explain to her that if she came in from that side there was a wall there. That held us up for quite awhile until in the end we just flipped it. We managed to change our planning to help her through it.
But that was an all good experience. I mean, I was a novice too and they couldn’t have made it easier on me.
How would you make the movie today if you wanted to do it again considering how the technology’s changed?
It would probably come down to our limitations again. If the budget had to be minuscule then you’d approach it one way and if the budget was unlimited – if you were looking at a Marvel thing – it’d be a relief to work with some of the big effects companies because you know they’re going to deliver.
But I’d probably put it entirely into like something like the Unreal Engine, combining the previz and the production process at the same time so you’re never really in previz, you’re always working with the finished quality. That to me feels like where things are headed.
They can’t do everything, and you have to rely on some conventional tools to fill the gaps, but in another few years I couldn’t fathom that they couldn’t handle most everything.
Would the same money buy you a lot more movie today?
Oh yeah, tons. The main problem with doing films that don’t have the larger budgets is that you don’t have the iterations that you do on a larger film. It ultimately comes down to and the amount of people that you can throw at it. You’ve got people that are talented but one or two people or ten people can’t do what thousands of people can do.
If you look at how long it still takes to render a frame from Jungle Book for instance, that’s out of reach for most people because it’s a massive commitment in terms of hardware and expense.
And if you can turn that into something that’s in real time on a laptop you can manage the entire shot and it’s always locked tight so that you’re effectively editing the film as you’re creating it.
You can suddenly stop the film, reach into the frame and move a building or change the sun, hit play again and you’re back live. It’s astonishing, and it’s a very hands on approach where you can kind of physically get in there and make the sets.
The film came out in 2004, you don’t have any major credits since then apart from 2012 short “Gumdrop”. What happened after “Sky Captain”?
It took so much time and effort and took so much out of me mentally and physically to make the film, it took at least 10 years total from start to finish. When I started it I knew what I was going to do every single day when I woke up. My day was filled. When the film ended it was almost like this black veil descended over me, and for the first time probably in my life, at least in a long while, I didn’t know what tomorrow was going to be.
I put everything into that one film and was disappointed in terms of the perceived success of it because of the people that worked on it. If I feel like I let anyone down it would be them. It was such a big leap of faith and I don’t think anybody will appreciate how these young kids were working all night, from early in the morning till three or four every night. They loved it, but the amount of dedication by this handful of people will probably never be appreciated except by me, ultimately.
So the whole process left me rethinking how to approach certain things. One of the first things that was offered to me right after that was John Carter, which Paramount had at the time, and I didn’t want to do it, for whatever reason I just wasn’t prepared for it. So I turned it down twice before they came back a third time and then I agreed to do it, thinking I’d be an idiot not to.
But unfortunately at the time, Sherry Lansing was running Paramount and she either retired or was let go [Lansing stepped down after Viacom – who’d owned Paramount since 1994 – split the company into two parts along film and TV lines]. An entirely new regime came in, and they weren’t keen on John Carter.
So after a year and a half’s worth of effort we put into that it was just gone. That was the same basic group who’d done Sky Captain and it was the first time I’d done something where I didn’t own the work.
Then I started another project at Dreamworks Animation, an adaptation of the Terry Pratchett novel ‘Truckers’. Jeffery Katzenberg was amenable to doing a live-action/animation version. But it was another six months to a year and it ultimately didn’t happen.
I was briefly working with Sam Raimi when he had an interest in doing “The Shadow”. I was also talking about ‘Doc Savage’, which is a comic I love. Then he lost interest in “The Shadow” so Sony lost interest in both projects.
So [each project] took a lot of time and energy and the work was really solid, so that’s when I decided I was just going to disappear, I’d just do my own stuff, and that’s basically what I’ve been doing. I’ve not been idle, I’ve been doing stuff that’s more about the experimentation [of “Sky Captain”].
That Hollywood tentpole world is fine and the people are generally speaking gracious and nice. But it comes with a lot of pitfalls and it just turned me off. The problem for me is that I’m incredibly slow, that’s why it took 10 years for “Sky Captain”. But I think I’m about ready. I think we’ll have something that’s not innovative for innovation’s sake but for the same reasons, something that takes advantage of the incredible tool set.
Do you find the endless procession of CGI smash-em-ups tiring?
They’re extremely successful and obviously audiences love these films. There’s probably a necessary sameness to them after a while, not that you can’t marvel at what they’re doing – no pun intended.
The work they’re doing is extraordinary, but it’s still being done by the handful of effects companies that do this routinely, and they’re not necessarily going to be the ones that bring a different look and feel to it.
Occasionally you get a film that takes some chances – “Guardians of the Galaxy” attempted some things that could have failed badly for them – so it does happen from time to time.
“Sky Captain and the World of Tomorrow” is out now on Blu-ray.