fxguide: this film has a reduced color palette with a lovely tonality to it. what was your approach in that reduced color palette? was that the intent from the outset with wardrobe and set design?they rebuilt the focus in *every* shot?? holy shit.
peter doyle: well bruno, the dp, was very keen to investigate some very specific color motifs. and the sort of general principle was to shoot with wide lenses typically from about 16-20mm. so basically have the angle of view of a wide lens but the depth of field of a long lens. so in fact, in all shots of the film we rebuilt the depth of focus.
fxguide: oh really?
peter doyle: it's probably not that apparent. it's not something necessarily you'd react to instantly. but if you look at the film you'll see that the foreground to background focus ratios is completely different and that most of the background is softened as opposed to blurred. in the process that will change the color and the saturation because as we all know as you blur an image you kind of reduce the dynamic range.
so by going into the film that way, in terms of just saying, well lets kinda mix the lenses that we're used to. because the grammar of cinema that we're kinda used to, if we see a wide shot we're kinda used to it being completely in focus, and if we see a long shot we're used to seeing the background out of focus and you could say that's almost a cue of what the shots are, because particularly now with modern lenses (bruno went with the torque s4's), their distortion is quite minimal between the short and the long. but the depth of focus is mathematically how the lenses are, you can't change that.
so mixing that around we ended up with a very interesting concept: and that's rather than changing the color and density of the film by in fact changing the density, we thought it would be far more attractive to let the contrast and the density and the color be changed based on the depth of focus.
because the thing is this is a cast driven film in terms of there's an ensemble. they almost always have three kids on screen and it's widescreen. that tends to dictate shooting things reasonably in focus. but if you look any of the shots you find the background typically far more difused and out of focus than in fact they are as captured by the lens. so that made for a very interesting concept because it meant we could grade for the foreground knowing that the background would in fact have a significantly reduced contrast.
fxguide: on a shot that had a virtual background - where you couldn't get mattes, did you have team of roto supporting you? how did you manage so well to define the foreground from the midground and the background?
peter doyle: we basically rotoscoped it. it's no secret these films take their time to grade. but no, its just myself and second colorist. we basically sit there and rotoscope it. you don't necessarily need the articulated mattes that you need to separate and do a true comp. we're talking far more generic kinds of shapes. it'd be more like the old days of working in a darkroom where you're dodging and burning. because the thing is, if you actually did a super-tight articulated matte it would actually draw attention to itself. the depth of focus we were doing was based far more on a plane of focus if you will.
it wasn't necessarily just drawing around the actor, it would be finding a vanishing point in the shot or finding some piece of set construction and drawing that. if you walk on set or walk around with the set painters they are in fact doing the same thing. if you ever walk on a set and look at a backdrop obviously it's completely shocking to see how little detail there actually is in there. but that's the whole trick of set painting and set design. in a bizarre way we're just taking that a little further. stuart our production designer and bruno would sit here and we'd define what should or should not be out of focus.
i didn't notice this effect when watching the film, so im interested in viewing it again. on the one hand this sounds like a bunch of creative tinkering designed to be interesting but ultimately not terribly useful or impactful. or it could be a practical way to overcome the limitations of lens technology and allow for a storytelling technique that hasn't previously been practical. since i was bored out of my mind for much of the film im not sure it's the latter, but i'll reserve judgement until i view it again.
if nothing else, this may be a technique that, although not necessary for an entire film, could be useful in certain situations - specifically when you want a wide-angle view but don't want deep depth of field. im interested to see if there's an emotional tie-in to violating that aspect of the visual language we've come to understand.
i'd wanted to do something similar with 'an exercise in vigilance,' but determined that it would be way to much work. however i didn't take the approach peter details here with finding a vanishing point. im not exactly sure how defining a plane works, especially when the plane only encompasses your actor(s). but perhaps the difference is simply working on a baselight with a 2k image in a viewing theater and having the luxury of taking six months to do your grading.
Posts
