Fellowship of Fans > News  > The Rings of Power and its photographic process

The Rings of Power and its photographic process

You know, I’m worried.

I mean, why not? Are you not sometimes worried? Are your friends and families not sometimes worried?

We’re all worried.

You’re worried. I’m worried.

Really, I am!

 

I worry for The Rings of Power. News of men falling down from the skies, of pixie-like Hobbits with anacrhonistically-English name, of Elves with Hobbit-like haircuts riding to battle in cool-looking but bordering-on-fussy armour, and that weird-looking Troll all have me worried that the showrunners are possibly preoccupied by making a “fantasy show” with fantasy visuals, and leaving behind the kind of quasi-historicist approach that made the live-action movies so successful.

Look, I like pre-Raphaelite paintings as much as the next man, but if every Lindon warrior Elf wears this kind of armour for every action scene, the cool factor is going to wear-off and the fanciful nature of the elaborate armour will instead sink in.

You’ll notice this criticism is largely based on what we see in the trailer. Key word there: what we see IN the trailer. As in, within the frames of the trailer. Not ON the frames of the trailer: I think too many people have become too preoccupied with the trailer’s photographic look in terms of saturation and crispness to really devote enough attention to the way the things within the frame look.

I think that shouldn’t be much cause for concern: trailers are notorious for kicking the saturation dial up to the maximum, and are systematically more vibrant-looking than the finished product. A good example is the trailer to The Desolation of Smaug – the varied imagery of the film (a personal favourite of mine) runs the gamut from lush to very muted – but its never anywhere near as vibrant as what its trailer would make you think. Compare this great closeup of Balin between trailer (left) and film (right):

The same, by the way, is true of the pictures from Vanity Fair’s first-look, which were clearly captured by a still photographer, on a large-format still camera, which captures much bigger and crisper images than any film camera could ever hope to capture. Notice, for instance, the extremly shallow depth of field on the shot of Galadriel above, typical of this kind of still photography, which is not something I except to see in the corresponding frames in the show. Those stills are not even in the show’s 2.39:1 aspect ratio.

But how can we expect the show to look? Its hard to tell, but at the moment let us devote just a few – oh, so very few! – words to their photographic process. According to the show’s IMDb page, its being shot with the Arri Alexa LF and Mini LF, used with Arri Signature Prime and Prime DNA Lenses. This has corroborated earlier by pictures of the filmmakers wielding said cameras, which had been used before by director JA Bayona and his director of photography Oscar Faura.

These are large-format cameras: a fancy-term which means they capture large-sized images that require less magnification to fit on a screen of any given size and therefore look better. Due to the inner workings of the sensor, they’re also less noisy and less susceptible to artefacts like halation. They output 4.5K ARRIRAW files. People will always tell you that resolution figures are misleading, and in a certain way they would be correct: the 4.5K figure is a drawer size: how full the drawer actually is would depend on how well its photographed – I can shoot with an Arri Alexa Mini but out-of-focus and still get 4.5K ARRIRAW files, but it sure as hell won’t resolve 4.5K. Actually, the full drawer size is by definition unattainable: the show is actually being shot in resolutions just short of 4K. [1]some 10-20% of the resolution is eaten-up by the built-in demosaic filter and the optical low-pass filter, and large format lenses like the ones used by the Alexa LF have certain optical aberrations … Continue reading

Surprisingly, IMDb also tells us that the show has a 4K intermediate, which I’m not entirely convinced with: even very recent effects-driven, high-budgeted films like Avengers: Endgame had a 2K digital intermediate. FYI, the term “Digital Intermediate” refers to a funnel through which the footage is filtered down: VFX are harder and more expensive to render the higher the resolution, and so they’re typically done in 2K and so the footage is funnelled down to this resolution. If what IMDb tells us is true, the show will be effectively lossless in that the effects will rendered at around the resolution that the footage resolves as it was captured, and the results should look quite crisp.

The lenses cited by IMDb are spherical lenses, which is something that was already apparent by the telltale lens flare exhibited by the shot of Valinor revealed early-on by Amazon. This is different to anamorphic lenses, which capture a a stretched-up image that’s then squeezed down to produce a widescreen framing without cropping: this creates a slight magnifying effect, but the footage itself looks a little bit less naturalistic because the degree of “squeeze” is not quite the same across all the planes of the image, producing oblong bokeh in the background [2]This makes rack-focus transitions rather tricky to perform smoothly in anamorphic, since in shifting the focal point, the film seems to squeeze and unsqueeze before the audience’s eyes. and oblong lens flare in the foreground.[3]Back in the analog days, anamorphic meant the footage could be shown at its native size without cropping, which again meant less magnification for a screen of any given size, resulting in a more … Continue reading

Anamorphic and Spherical lens photography: above left, extreme anamorphic lens flare (seen as streaks of light across the full width of the frame) from Apocalypse Now, as compared to spherical lens flare (seen as a series of short streaks and round artefacts of coloured light) from The Two Towers. Below, left: oblong anamorphic bokeh from Braveheart, seen as candlelight stretching upwards, while a similar out-of-focus candlelight behind Saruman in The Fellowship of the Ring appears round.

Its usually at around this point that the filmstock purists come crying bloody murder for the fact that the show is being shot digitally and not on filmstock. I’ve argued before that we shouldn’t make a religion out of practical effects, and the same is all the more true of analog cinematography. Not that I have anything against using filmstock: some of my favourite films (photographically-speaking or otherwise) were captured on film: see three prominent examples above. But many of my favourite films were also shot digitally, and I think to cling to filmstock, at least to the extent that some people in the fan community are, rings to me of the reactionary, which is something that I’m not sure an artform as inherently technological as cinema can abide.

This is especially true of TV and streaming, where a lot of footage is generated across a long period of time which makes shooting filmstock expensive, and also cumbersome from the perspective of the pipeline: since you don’t need to reload the camera with a filmstock magazine after every couple of takes, and since you can shoot later into the evening, digital generates more footage more quickly, and you can take it all but immediately into the colour grading suite, you can composite shots much more easily (because you don’t have to match disparate grain structures), crop shots somewhat to achieve a slightly more desirable framing, etceteras. “ah, exactly!” I hear you cry, “its the rigour of the limitations imposed by shooting on film makes for better filmmaking.” This is such a reductive argument, its tantamount to saying that cutting on a steenbeck makes for better editing, which is of course something many people have said but its not true and certainly not practical: no filmmakers in their right mind will willingly make his or her own job harder than it needs to be, and just because necessity can be the mother of invention doesn’t mean that to deliberately set obstacles for oneself makes for better art.

“But, surely, this will take away the grit from the show?” Oh, really? Some of the grittiest action films in recent memory were shot digitally: Mad Max: Fury Road comes to mind, but so does perhaps the grittiest of all action films, Apocalypto (see below).  They’re not even films that particularly seek to disguise the digital acquisition by desaturating the footage or applying grain and halation filters and whatnot: they look digital, and they’re crisp, clear and vibrant, but – and this the important part – they’re gritty because they tell a grim story which contains grit within the frame, not on the frame. Another film that comes to mind is Skyfall, brilliantly photographed by Sir Roger Deakins, who had all-but sworn-off film, and doesn’t to my eyes seem particularly occupied with making his digital cinematography look particularly film-like, but who nevertheless captures arresting images. Again, the grit of that film (and others like it) is within the frame, not on it.

Dear reader, let me introduce you to my friend, “gritty”. Gritty, this is…um… sorry, I didn’t catch your name, dear reader…?

“But then, the extreme clarity and crispness are sure to give the footage an ‘uncanny valley’ look à la ‘soap opera’ effect, right?” Well, not really, no. The “soap opera effect” isn’t caused by high resolution, but rather by shooting at a higher frame-rate, which is not the case here. As for it looking overly crisp, I think that’s special pleading: When Sir David Lean was shooting films in 65mm in the 1960s no-one criticized the format for being too crisp, and nowadays when Christopher Nolan and Quentin Tarantino shoot on VistaVision, 65mm (spherical and Ultra Panavision 70) and bloody IMAX,[4]VistaVision is to 35mm what landscape mode on your phone is to portrait mode, and the same is true of 65mm and IMAX. Ultra Panvision 70 is an extreme anamorphic format used with 65mm film which … Continue reading their tenacity to shoot in these cumbersome formats – precisely for their superlative clarity and sharpness – is lauded as an act of great showmanship. Its only with digital cameras that this argument is dredged up: the Alexa LF resolves less information than Nolan’s IMAX footage and shouldn’t create any more of an uncanny valley effect that his footage does.

“Exactly! They should be shooting on film because it resolves more detail!” Well, not really. For one thing, this is a streaming series, and on a TV screen, its impossible to discern the difference in resolutions exceeding 5K and, in most home theater setups, even above 3K. What’s more, the argument for the resolving power of filmstock had been greatly exaggerated, in part because people conflate the resolution in which filmstock is scanned (which, for technical reasons, is always oversampled). Yes, its common to scan IMAX footage in ~12K, but that doesn’t mean it actually resolves 12K. Actually, it perforce means that it resolves less, because as I’ve said, the scanning resolution is and must be oversampled: I’ve seen camera tests where IMAX footage resolved less than the LF’s big-brother, the 6.5K Alexa 65, and even marginally less than the non-large-format 6K RED Dragon. This idea that film is somehow ungodly resolute is a canard.[5]To properly capture the grain structure and to avoid aliasing, film should theoretically be scanned at 2.5 times the resolution it resolves: Well-exposed, 100ASA Super-35mm footage theoretically … Continue reading

“Ah, but using filmstock is an aesthetic choice!” Sure, but so is shooting on 8mm film; so is black-and-white photography. The fact of the matter is, 99% of filmmakers in 99% of films and shows will choose to use the most naturalistic-looking format they can afford and manage, and the aesthetic choices are then made by subtle tweaks to the look of the format through lighting and grading, rather than by digging-up some 1950s CinemaScope camera in order to use their infamous “anamorphic mumps” abberation to make some artistic statement. [6]Rest assured, too, that if you want to make your RED Monstro footage look like 35mm, the technology exists that will make it indiscernible from it: even Quentin Tarantino, ever the analog-zealot, had … Continue reading

Honestly, if we want the show to have something of the naturalism that the films had and strived for, would we not want it to have naturalistic-seeming photography, without overly-muted and stylized colours or excessive gauziness (what I call the “shot-on-parchment” look [7]There are shades of this look in The Fellowship of the Ring, The Two Towers and – in an attempt to bridge the gap to those films, The Battle of the Five Armies; but perhaps the film I most … Continue reading) and without artefacts like excessive dye clouds? I know that’s what I would like. Perhaps not quite the overly-saturated golden hue of the Lindon trees from the trailer, but goddamit, I like COLOUR in my cinematography!

And, really, I’m much more interested in the actual mise en scene: the framing, the blocking of the camera moves, the composition of the shots, the cutting from one angle to the other, than I am with the look of the frame itself. I like quite a few films in spite of their look: I’ve never been crazy for the “picture book” softness of some of The Fellowship of the Ring, but I’m much more interested in the exuberence of Jackson’s “flyovers”, the unobtrusiveness of his oners, the starkness of his closeups, then I am with nitpicking the minutiae of the colour saturation and grain structure of his shots. Another example is this fantastic oner in The Battle of the Five Armies: the look of the shot is not particularly ingratiating – a bit too desaturated and airbrushed for my tastes – but the way the shot keeps on going, with the coordination of all the people, and the way the movements are motivated by movement in the frame to the extent that the longness of the take never calls attention to itself. That, ladies and gentlemen, is directing. Perhaps my favourite example, in all of cinema, of how cutting from one angle to another tells us something visually is when we cut to this wideshot: you cut to that shot of Sam against the slope, and the wideness of the framing (all the more accentuated from having cut to it from a closeup) conveys just how insignificant and helpless Sam feels in the face of his predicament: Mon Dieu, c’est Cinéma!

So, if you want to worry, be my guest and join the club (NB: We have cookies!) but, for the meanwhile, worry more about what’s in the frame, and less about the frame itself!

References

References
1 some 10-20% of the resolution is eaten-up by the built-in demosaic filter and the optical low-pass filter, and large format lenses like the ones used by the Alexa LF have certain optical aberrations that also reduce the drawer size.
2 This makes rack-focus transitions rather tricky to perform smoothly in anamorphic, since in shifting the focal point, the film seems to squeeze and unsqueeze before the audience’s eyes.
3 Back in the analog days, anamorphic meant the footage could be shown at its native size without cropping, which again meant less magnification for a screen of any given size, resulting in a more resolved-looking image. This in spite of the fact that anamorphic lenses tend to be a bit less sharp than their spherical equivalents. Because of the lack of grain, digital is more amenable to magnification (that’s why a picture shot at around 960p like Attack of the Clones could be shown in IMAX whereas 16mm films cannot) and therefore this effect is relatively negligible and so filmmakers are gravitating back to spherical lenses which produce a more naturalistic look, are less cumbersome with rack-focus transitions and are easier to integrate VFX into.
4 VistaVision is to 35mm what landscape mode on your phone is to portrait mode, and the same is true of 65mm and IMAX. Ultra Panvision 70 is an extreme anamorphic format used with 65mm film which produces an aspect ratio of 2.76:1, known to audiences from The Hateful Eight and, prior to that, from Ben Hur. There’s also an anamorphic VistaVision format called Super Technirama, used by Kubrick in Spartacus. All of these, along with the original 2.55:1 CinemaScope, were born to emulate the cumbersome Cinerama format, which required four rolls of 35mm film to record one ~2.55:1 pin-sharp, wide panorama, projected over an immersive, curved screen (at 27fps, no less!) with the surround sound occupying the fourth reel. Cf. How The West Was Won.
5 To properly capture the grain structure and to avoid aliasing, film should theoretically be scanned at 2.5 times the resolution it resolves: Well-exposed, 100ASA Super-35mm footage theoretically resolves 3.2K, and so some films like The Wizard of Oz were scanned at 8K. Naturally, most real-life photography doesn’t produce anywhere near that much moiré and so one can get away with scanning the footage just in moderate excess of its resolving power: 35mm is usually scanned at 4K and Nolan’s IMAX footage is scanned at 8K. Now, since Super-35mm is 24.96mm across and IMAX is around 69.6mm across, it should resolve 8.9K, but IMAX lenses have greater aberrations than 35mm cine-optics and suffer from issues like dye bleed. Plus, these figures count the grains, which are technically artefacts. If we look purely at picture information, 35mm resolves around 2.5K and and IMAX resolves around 5K. Tellingly, except for a few shots in The Dark Knight, Nolan’s VFX are rendered at around 5K. Of course, the grain themselves somewhat contribute to the perception of sharpness and detail, because we see them as pinpoints in sharp-focus dotted across the frame, and because they’re randomly distributed across each frame, so our mind “averages” the amount of detail on every couple of adjacent frames, and so the distribution of the grain means that even in static shots there’s more information across a couple of frames than any individual frame. These effects are somewhat approximated by the demosaic on digital cameras, and since they diminish with format size, they’re not very relevant to the sizes we’re discussing.
6 Rest assured, too, that if you want to make your RED Monstro footage look like 35mm, the technology exists that will make it indiscernible from it: even Quentin Tarantino, ever the analog-zealot, had to be told that Apocalypto was shot digitally: he couldn’t tell from having seen it printed back unto film.
7 There are shades of this look in The Fellowship of the Ring, The Two Towers and – in an attempt to bridge the gap to those films, The Battle of the Five Armies; but perhaps the film I most associate with that look is Harry Potter and The Half-Blood Prince (which I’m very fond of, by the way).
Share
Chen

Historian and perpetual Wagnerian, I had discovered the Lord of the Rings along with Tolkien’s other, multifarious writings after the release of The Fellowship of the Ring in 2001. As an avid filmgoer and writer, I take a particular interest in adaptations of Tolkien’s works – past, present and future, realized or otherwise – and participate with Fellowship’s podcasts in that capacity, researching and discussing the Amazon show.

1 Comment

  • Gonff
    May 9, 2022 at 5:18 pm

    I very much enjoyed this write-up. You clearly know your stuff when it comes to the process of image capture and processing and I couldn’t agree more… The camera is just a tool. It’s what’s within the frame that counts.

    After the main trailer was released during the Superbowl, I completely recalibrated my expectations for the show. I thought that they would have respected the scholars and legacy of Tolkien a lot more than they appear to have done. Now it just looks like big budget cash-in like the latest Jurassic World Films. Oh well, opportunity missed. I will watch the first episode and judge it then.

    Also, I think there’s something wrong with your comments section… when I type, I can’t see my text unless I highlight it.

Add Comment