Big action movies – particularly superhero films – are these days often enough in 3D that Fantastic Four being a 2D-only release is news. But these days me and my friends usually specifically go for 2D showings. It’s just not worth the risk of mild eyestrain – most films are converted in post in the most obvious way possible and don’t particularly benefit from the treatment anyway.
Ant-Man could have been different, though. It’s the first big Hollywood size-changing film to be presented in modern 3D. There’s scope to do interesting work there: to show small sets as impossibly vast, to really sell the change in scale. At best it might have been as good a use of 3D as the Doctor Who 50th anniversary; at the worst it would be interesting to see what they’ve done with it.
Ant-Man was shot in 2D, which is an immediate mark against. My short list of films that strongly benefitted from being in 3D (that Doctor Who special, Tron Legacy and Gravity) has two that were shot in 3D. But Marvel haven’t used them once. This is not exactly a budget thing (yes, they’re expensive, but so is not using them), so much as practical logistics dealing with two cameras rigged together into an unwieldy contraption. And while this is difficult enough at the human scale, doing it at the macrophotography level (there was an entire macro unit on to get shots of insects and such forth) would have been proper bleeding-edge stuff. So instead Ant-Man fixed it in post: using the by-now standard process where artists designate depths of various areas, and tools generate the appropriately shifted images.
Unfortunately, Ant-Man‘s use of 3D is an unimaginative afterthought. The framing of shots works against it – a frequent tic is having two-shots with an out of focus back of the head overlapping the side of a frame, something that looks really nasty in 3D, and most have been a nightmare for post-conversion (which likes nice sharp edges). Action is kept mostly on the plane, which is good, but combined with a very shallow depth of field it feels actively hostile to the roving eye in a way the best 3D films avoid.
There are a few actively botched shots, including a woman in a crowd scene who has been placed much farther back on the screen than she should be, making her appear to be about 10 feet tall, compared to Scott. (If you want to watch out for it, it’s just after he gets sacked.) If this were more consistently applied, it might be a joke, making the future Ant-Man appear small even before he gets the suit. Some glitches are hard to avoid: stereo is not kind to high-contrast environments like the Yellowjacket vault: the contrast of Hope’s hair and those bright-light backgrounds look wrong. (Oddly this is of those areas where having 3D source material wouldn’t necessarily help – Tron needed extensive fixups to match up reflections and glares between eyes, using software I helped work on.)
The first size-shifting action sequence – the one with the bathtub – is a stunningly well-designed sequence. It makes its points – that the world is very different at scale, and that Ant-Man is not a precarious bug, but a little bullet – with a real dynamism and integration of elements, rather than the oversize props and unconvincing inserts we are used to from pre-CGI size-changing films. What makes it work is the discipline involved in keeping the camera at a small scale for the entire sequence, with cutaways kept brief.
But this discipline is lost for the rest of the film. The middle action sequences (the escape, the fight with the Falcon, and the heist) tend instead to adopt a compromise framing, where Ant-Man is shown quite small on the frame, but not too small that you can’t see Paul Rudd act. The elements within this that work best (the pipes scene, the servers) are those isolated from the human scale entirely: the fights against proper humans are confused – we are watching something incredible happening without really understanding its rules. The final boss fight returns a little to form. Because Ant-Man is fighting someone his own sizes, the camera is allowed to get close to them, and to show off the massive vista of Cassie’s bedroom: the cutaways are excellently used for context and laughs. But all this has been undermined by muddy use elsewhere in the film.
This film had a reason to be in 3D, but what interesting things it does with it are almost by accident. And if a film with this much potential – and a film that was always going to struggle at providing any reason for anyone to see it – didn’t manage to embrace it, is there any point pretending big studios see the technology other than a cash-in? For the time being, I’m going back to 2D.