Focus pullers, traditionally, don’t complain much. Nobody’s a mistake-proof machine, but saying that out loud represents an admittance of fallibility that many people prefer to avoid, especially if the ability to sleep indoors and eat food is based entirely on one’s ability to judge distances and manipulate a lens appropriately. It therefore, takes quite a bit of persuasion or beer, before most of them will accept that the new world of huge resolutions, huge sensors and stygian on-set gloom can be a little trying, especially in a world where gimbals, the Alexa LF Mini and rushed schedules with no time for blocking all exist at once.
Or, to put the situation more boldly, no, it isn’t really possible to keep these sorts of pictures in focus down to the maximum capability of the sensor. We’re not even really trying to do that anymore, at least not with any expectation of actually achieving it. Tellingly, not one of the people I spoke to in the preparation of this article was willing to go on record as saying this. The knowledge that we really can’t fully satisfy 6K or 8K sensors, most of the time, puts the lie to an awful lot of convenient marketing half-truths about how much resolution matters. Who’s shooting 8K? Practically nobody, not in any real sense. Perhaps the actor’s leftmost eyelash occasionally drifts through the zone of focus in a particularly sharp region of a particularly sharp lens, but on anything other than a sunlit day exterior on the best glass in the world, real 8K pictures are tough to come by, regardless of the number of photosites on the sensor.
The thing is, this is good. It’s not a failure. It’s a triumph; it means that the resolution of the recording is no longer the limiting factor of what we can see. Focus pulling is one limiting factor. Lenses are another. But surely, if we have so much spare resolution we can happily use some of it to record the characteristics of the lens — isn’t that what we’ve always wanted? It’s almost a return to analogue, in that, sure, the image is being electronically captured and recorded, but the impact of that is so microscopically tiny, at least in terms of resolution, that it has practically no effect on the images we can create. If nothing else, it makes the softness imposed by antialiasing filters and Bayer demosaicing almost irrelevant.
The post production chain
In early 2020, downstream processing in post production or distribution will very often impose technical limitations on the sharpness of what can be seen, just as 35mm film distribution did. With cameras beyond 4K, however, cinematographers are now in a position to make sharpness a creative concern, not a technical one, within reasonable limits. Back in the day, a film shot super-35 with ancient lenses on ISO 500 stock had vastly lower resolution than something shot in widescreen on ISO 50 stock and spherical lenses. That was fine. Nobody complained and nobody should complain now if the quality control department receives an 8K or even 4K master that doesn’t target pixel-level sharpness on every shot. It’s impossible, most of the time, but it’s not even always desirable.
Cameras with enough pixels for all this to matter used to be expensive. Now they’re mainstream, so everyone has the choice. Probably we should all realise that, yes, we pursued these resolution increases so that we would have it in case we needed it and we don’t have to use it if we don’t want to.