Alita: Battle Angel

Alita 1

It has been quite a while since I have written a post on 3D. Alita: Battle Angel (2019) marks a type of anniversary for me. It’s almost 10 years since Avatar’s (2009) release, when I first became serious about studying 3D cinema. What’s fascinating about Alita is that, for all its advanced visual effects, it closely follows a stereoscopic 3D formula established by Hollywood almost a decade ago. This is not entirely surprising considering director Robert Rodriguez and producers James Cameron and Jon Landau have a long history working in digital 3D.

Alita is also a film that shows Hollywood’s attempts to move forward in the area of representation whilst keeping a foot firmly in the safe narrative strategies that maintain box office appeal. While Alita has one of the few female action hero leads of the last decade, and while the casting has been hailed as diverse, this is still Hollywood holding on to storytelling of the past: a heterosexual love story forms the core of the emotional entanglement; beyond Alita women of colour are given bit parts in the form of a nurse and women-to-be-assaulted-at-night; one token women is added to the gang of bounty hunter bad asses (no speaking role); the first significant man of colour to be seen, Vector, is (of course) the baddie.

alita 2.jpg

It’s within this framework that the stereoscopic effects push at the boundaries of visual imagery but are closely tethered to the narrative so that safe storytelling can take place. As the film opens on mid-shots of Dr Ido amongst the debris of a massive junkyard underneath the sky city Zalem, there is carefully placed fore, mid and background depth. This is emphasised in the shot when he discovers the head and torso of a cyborg: the small screen with translucent data that he holds up to scan the cyborg provides the midpoint in a three stage depth set-up. The next scene, inside Ido’s laboratory mixes a steam punk aesthetic of antiquated machinery and equipment with numerous translucent screens that emphasise the spatial possibilities of data. 3D is used here to emphasise the configurations of a dystopian future world that we are gently invited in to.

When the cyborg, Alita, wakes up in the next scene she is surrounded by comforting blankets in a snugly bed. Striking close-ups on Alita’s face, drawing attention to her large eyes, and then her robotic arm as it stretches horizontally across the screen give us the sense of the uncanny valley. However, the stereoscopic depth has placed us in close intimate contact with her. The 3D allows us an emotional connection as we come to terms with Alita’s difference. Not just difference in that Alita is a cyborg but difference in that we have a female protagonist in this richly created science fiction world.

alita 3.jpg

Downstairs, in Ido’s kitchen the over the shoulder two-shots between Ido and Alita are arranged to display further emotional connections – joy in Alita’s face as she tastes an orange for the first time; tears in her eyes as she realises she doesn’t know who she is; Ido’s empathy with both these moments. The stereoscopic depth enhances the placement of these two characters but there is also a strong sense of ontological difference. This is not simply at the narrative level: between a cyborg and a human. Rather it is at the level of the image. While Alita’s face and body are rendered in fine detail, she is still visibly of digital construction and not of the same visual constitution as the photographed actor playing Ido. This is emphasised by the subtle depth relations that make the difference in detail between the two apparent. Further to this, while stereoscopic depth emphasises where they are placed in relation to one another, the use of shallow focus suggests the characters have their own, particular, space in this relationship.

The first third of the film is typical of these moments and thus typical of digital Hollywood 3D. There are extensive depth relations in positive parallax space but objects in the foreground – mostly static – only protrude very gently into negative parallax space. There are almost no objects traversing the z-axis from foreground to background or vice versa. Stereoscopic depth is used for the creation and emphasis on relationships and spatial placement but is not meant to draw attention to itself or be seen as assaulting the audience. One of the few exceptions is when Alita and love interest, Hugo, are on top of a building in Iron City. There is one vertiginous shot where they/we look down the side of the building to the street far below at the same time as an aerial points towards the audience. It is the slightest teaser of the more dynamic depth relations that will come later in the film; a temperate reminder that there is a reason we have paid the 3D surcharge.

alita 4

A little later the first battle between Alita and other cyborgs takes place as she and Ido are confronted in an alleyway. There is a similar ontological difference between the CGI cyborgs and the human Ido. It is the moment when the film confronts us with the alien difference of Alita – with her in built violent responses – while also asking us to invest in her powerful momentum. Just before Alita attacks the cyborg Grewishka, she performs a high kick that seems aimed right at the audience as she traverses negative parallax space. This is followed by Alita’s movement forward that leads to an intense close-up on one of her eyes: a cinematographic feat only stereoscopically possible with advanced digital compositing. It is the stereoscopic equivalent of the narrative turning point: a ramping up of depth intensity which sets a new norm for depth intensity and use of negative parallax space for the rest of the film. It is no coincidence that it comes at exactly the same moment as the narrative’s first turning point.

As the film progresses Alita’s strength increases and stereoscopic effects become more intense. In the bar fight, stereoscopic debris occurs as glass shatters towards the audience and splintering wood shoots into negative parallax space. Notably, the former takes place in slow motion, announcing the moment at which the film has sufficiently gained the trust of the audience to overtly show off its spectacular manifestations. After the initial bar fight, Ido enters and manages to get the fighters to calm down and remain still. Like the choreography of a complicated dance momentarily paused, the stereoscopic depth allows the spatial relationships of the characters to be seen. It is only a short pauses and Grewishka soon enters the scene. The audience is now primed for full use of the z-axis and this is granted when Grewishka crashes through the floor into the ‘underworld,’ allowing a fight along the z-axis to ensue between him and Alita. His mechanical tendrils frequently shoot directly towards the audience as well as away from them, often in slow motion so the full effect of his assaultive power is clear.


In line with the flow in and out of intensity in the narrative’s dramatic action, the use of stereoscopic depth fields also strengthens and relaxes. There is the ubiquitous use of an underwater scene when Alita jumps down into a lake where a URM spacecrafts from Mars crashed. In this moment, objects surround us in a tranquil manner while Alita appears calm. Soon after, action and stereoscopic effects gear up again.

When the second, climatic, Motor Ball battle occurs, it is much more kinaesthetically charged. Alita has mastered the use of a new cyborg body and we are invited along for the ride. Shots are set up in a phantom ride aesthetic where we feel that we are propelled into the fast moving scene. When the characters break out of the Motor Ball arena and along a series of pipes above the city, the perforated pipes gush water straight at the audience so that we experience the same momentum and counter force as the characters do.

Combined, stereoscopic effects and a narrative arc that takes Alita from broken mechanical parts to a dynamic action hero allow the audience to fully understand the irony when Alita states “and I’m just an insignificant girl” before slaying Grewishka. What is significant about Alita is that she has paved the way for a reinvigoration of the female action figure. However, this isn’t without cost.

When Alita is helped by Ido to enter her new cyborg body, one made from URM material, he notes that the cyborg body adjusts to match Alita. Ido says: “her shell is reconfiguring to the subconscious image of herself.” Significantly, this subconscious image, which becomes materialised, is the impossibly thin yet curvaceous body: delicately thin limbs and waist combined with substantial breasts and buttocks. So far, this body has only been possible in graphic illustrations or via substantial photo shop. I’ve written previously about the problems photographers have encountered when trying to manipulate bodies to meet this ideal at the same time as using stereoscopic depth. Alita is able to achieve it only via the use of a cyborg body created in performance capture.

alita 5

In this way, Alita sums up the possibilities of combining stereoscopy and advanced visual effects to meet both narrative and aesthetic future-facing aims. However, Hollywood has a way to go before it can truly shine light on the potential for diverse bodies and new action heroes.


Stereoscopic Star Wars: The Force Awakens



I am the worst of the 30-something cliched Star Wars fans. Having fallen in love with the original three as a small child and having hated the first three when they were released in digital splendor, I have just enough experience to be highly opinionated but haven’t enough knowledge to hold my own in a proper Star Wars debate. There are plenty of us putting our opinion out there online and so I won’t pretend I can add much to analysis of Star Wars: The Force Awakens’ plot and characters even when I did like it a lot and fell in love with Rey and what she adds to a male-dominated global Hollywood. Instead, I’ll do what I usually do and write about what was going on with the stereoscopy in the film.


The Force Awakens should have been a stereographer’s dream film. It has everything you need to showcase the latest advancements in digital 3D technology: plentiful battle scenes allowing plot motivated material flying at you in the auditorium; vast science fiction landscapes to demonstrate positive parallax receding depth; large space crafts that produce the cavernous interiors that 3D depicts so well; phantom rides as the camera moves rapidly into scenes, made more exhilarating by stereoscopy; and various translucent telecommunication scenes that show the way different depth planes can layer on top of one another. But it was a post-production conversion to 3D and at times it shows.


For starters, take the opening title – “A long time ago in a galaxy far away.” The neon blue letters on a black screen are faithful to the original but whereas the original had a slight analogue shake that softened them, these letters look they have been added using the Basic Title function in Final Cut. Although they are aligned, the contrast between the letters and the background sit uneasily in stereoscopic depth, not seeming to rest on a particular plane but rather floating somewhat out of synch. Following on, the famous explanatory text then nicely scrolls backwards into positive parallax space but a strong decision about the starscape behind it hasn’t been made.There is neither the textural fabric of stars (not stereoscopically accurate due to our inability to perceive stereoscopic depth when looking at objects that far away but visually beautiful) that have appeared in Gravity and The Martian nor is there a straightforward flat starscape at a distance. Instead, there is a rather weak dusting of twinkling lights that does nothing to engage the viewer as the starscape transitions into the opening shot of the planet Jakku. Complaining about this probably makes me seem very pedantic but these are the small details that a good stereographer or cinematographer with an eye for stereoscopy would work on.


Shortly after, when Poe Dameron and Lor San Tekka step outside to watch the First Order ships arrive, the rack focus technique is used to shift Poe out of focus in the foreground and bring Lor into focus in the background. Racking focus like this is common across contemporary Hollywood but it creates a more striking visual effect when performed in 3D cinema. It suddenly becomes more emphatic and there is a noticeable transition as two separate parts of the screen space seem to transform simultaneously. In the 2D version it is akin to our eyes gently shifting from one area to another whereas in the 3D version it seems like the visual world undergoes a significant change. In this way it is a little like the dolly zoom that was made famous in Jaws, a dramatic technique that adds tension and suspense to the scene. This would be fine if the scene needed the extra emphasis but it seems as if the shot was designed for the 2D version and was only meant to be a subtle shift from one character to the other. Throughout the film other shifts in focus occur that take on their own visual emphasis in 3D but seem to do so against the flow of the aesthetic and editing rather than with it.


Following a trend initiated in Avatar and epitomized in The Hobbit trilogy, roving cameras are used that sweep and soar with the characters as well as dart towards and away from them. Their use goes against orthodox approaches to stereoscopic filmmaking that prefer static shots and long takes so that the viewer’s eyes have time to take in extended depth relations. Nonetheless they fit a contemporary Hollywood style and public expectation for dramatic action scenes so it is unsurprising that the majority of 3D blockbusters use them. While some Hollywood 3D films are able to construct these shots so that they interact with the spatial configurations of objects in the back, mid and foreground, there is little evidence that this has taken place in The Force Awakens. Instead, stereoscopy is only able to enhance depth relations in a very limited way in these scenes.


At the same time, post-production conversion technology has come a long way in the last few years and when there is ample material to work with, there is still the ability to produce scenes that are visually more dynamic than their 2D counterparts. In The Force Awakens this occurs most often in the close-ups on characters when the cameras finally rest and there are relatively static shots. For example, when Finn decides to defect from the First Order, every bead of sweat and furrow in his brow is clearly and intensely visible, an aspect matched during the close-up of Poe when he is tortured by Kylo Ren. In another relatively static shot, shortly after being introduced to Rey, we see her slide down a vast sand dune and the stereoscopy helps emphasise how small she is amongst the great trench of sand. While the film has been praised for not relying too heavily on CGI, one aspect that works well with the stereoscopy is the depiction of telecommunication scenes and holography. When the final piece of the holographic map is inserted and Hans Solo walks through the translucent data fields, it feels like we walk through it with him.


As much as the telecommunications screens are a nod to our digital present, the film does a good job of pointing to the past. Surprisingly, the cheesy wipes and other transitions that haven’t been seen much in recent years work fine stereoscopically and don’t upset our visual sense of space and place. There are also occasional nods (whether intentional or not) to the famous moment in twentieth century 3D when Grace Kelly reaches her hand out directly towards the audience in Dial M for Murder. It occurs most often when characters reach for their lightsaber, a shot set-up that exists in the original films but becomes emphatic as it seems that they really do reach towards us at the same time we know they have no idea we are watching them.
There are hints, then, of how stereoscopy can add to the visual worlds created in the Star Wars franchise and in many ways this is what is most frustrating as it shows what could be achieved if a proper stereoscopy team were brought on board rather than merely handing over to the post-production team at the end. Whether or not Rian Johnson, director of the next two films, will take a different approach is yet to be seen but considering he wrote in 2012 that he would never shoot in 3D, I’m not too optimistic.

The 3D Cinema Book is out

I’m really pleased to say that my book on 3D Cinema has just been published:



3D Cinema: Optical Illusions and Tactile Experiences
by Miriam Ross
Hardcover 240 pages ISBN 9781137378569

more info available on this page

Man of Steel 3D


Since its release, Man of Steel (2013) has divided critics between elation at its reinvention of the popular Superman myth and disappointment that it has failed to deliver on its much hyped premise. Indiewire has a very good in-depth breakdown of the best and worst of what the film has to offer. To the site’s credit, its description of the film’s family theme almost made me revaluate what I had seen and it moved me to believe I had witnessed a strong interpersonal drama. However, on reading back through my notes I was reminded of how hard it was to care for or engage with any of the characters in the film. I am not enough of an expert in superhero franchises to provide a definitive account of the strengths and weaknesses of the way Man of Steel developed its story but I will try to make some points about how the film’s stereoscopy had an effect on making that development happen.

In an approach that seems to be trying to offset the problem the film has in eliciting identification between characters and audience, various shots bring us extremely close to its central figures. The opening shots, when Clark/Kal-El’s mother Lara is giving birth to her son, are filmed in extreme close-up. There are small glimpses of her face and hand where the texture of the image becomes haptic and indistinguishable from any sense of deep space around them. There isn’t a profound sense of stereoscopic depth but the haptic nature of the images, combined with the gentle stereoscopic effects mean that the screen plane is impossible to determine. A similar set-up is returned to at various points in the film. When a young Clark/Kal-El locks himself in a cupboard at school, there are tight close-ups on his face. In this case his face is more easily discernible but the sense of closeness remains. The same happens in one of the later battles scenes when the enemy, Zod, is shown within a very tight close up. Only his eyes and nose are visible and they seem to bulge into the audience’s space. In a later shot, the same happens with a close-up of Clark/Kal-El. By using stereoscopic depth, the shots are able to suggest a touchable, palpable quality to the characters but this provokes intrigue rather than sympathetic engagement.


The way in which stereoscopic effects provide intrigue rather than involvement in the film continues throughout Man of Steel. During initial scenes, shot on Krypton, lens flare occurs. Unlike its recent use in Star Trek: Into Darkness (2013), where it provided multiple planes of depth, in Man of Steel it delivers a scrim type of screen that separates the area of the auditorium from the action that appears in positive parallax, behind the traditional plane of the screen. A similar type of set-up is used throughout the film although there are some shots, most commonly those that place characters in the foreground, where body parts bulge slightly into the auditorium. Throughout the film, this is about the limit of negative parallax that is used. When Jor-El, Clark/Kal-El’s father, dives underwater to escape enemies chasing him on Krypton, there is a hint at the thick space created in stereoscopic underwater scenes but there is little sense that the space breaches and enters into the viewer’s physical space. The particles and bubbles that are seen are more often in positive parallax, far away from the viewer’s body. This set up is repeated when Clark/Kal-El  also finds himself underwater in a later sequence. Shortly after, when Lara and Jor-El argue about what to do with their baby, an intimate shared space is created between them by the use of dust particles crossing different depth planes. However, these dust particles rarely occupy an overt place in negative parallax and so the intimate space is not shared with the audience. When Clark/Kal-El embarks on a rescue mission to save men trapped on a burning oil rig, embers float around the screen space, but again not in negative parallax.


Although it might seem that these points are mundane, the reason I am making them is that these moments reference classic set-ups now seen repeatedly in digital 3D films but their full stereoscopic potential is not realised. Another classic set up for stereoscopic films is initiated when the school bus carrying Clark/Kal-El swerves and goes over a bridge. As it crashes into the water, liquid is thrown up and towards the audience but does not fully enter negative parallax space. In a final nod to the use of floating particles in stereoscopic films, there is a brief scene in a snow storm when Clark/Kal-El walks out of a bar. The decision for the snowflakes to stay mainly in positive parallax seems, by this point, to be an indication that the film is determined not to follow in the footsteps of the trends initiated in other 3D films. In a similar way to the lens flare, these moments make it seem as if a screen is placed between audience and action and that a thick, textured action space lies behind it but not in the auditorium.

It is only much later in the film, when the actions sequences accelerate, that negative parallax is employed more frequently. When the battle between the enemy, Zod, Clark/Kal-El and the humans intensifies, there is a tendency for the typical stereoscopic debris found in battle – exploded material, fragments of building, crushed rock, smoke, embers and fire – to be projected into the auditorium. These effects are combined with intense sound effects, fast camera movement and an overwhelming visual image that is relentless for the last quarter of the film. A signal that the film is going to present itself more forcefully towards the audience comes just before the shift into this final action quarter, when Clark/Kal-El has just finished experimenting with his flying abilities. After crashing back onto the ground, his hand reaches out over a piece of rock and seems to extend towards the audience. This shot is almost repeated at a later point when Zod’s hand crushes dust in a tight close up that extends towards the audience. In each case, the film seems to be trying to tell the audience that they are imbricated within the personal drama of these central characters. However, in a similar way to the close ups, it seems as if the visual field is desperately trying to create a relationship between audience and characters that the film’s plot has failed to deliver.

Fusion x64 TIFF Fileiron_manPin_Art_Cover1

What was most interesting for me was the update to data communication screens commonly seen in contemporary science fiction 3D films. Various films such as Avatar (2009) and The Avengers (2012) have used stereoscopic effects to show off translucent telecommunication screens. They seem to suggest that we are only a generation away from tactile screens that, unlike the hard bodied touch screens we currently use, will be transparent and weightless. The first film to add a variation was Prometheus (2012) which included screens that formed images in a way not dissimilar to the pin-sculptures that were popular in the 1990s. In Man of Steel a similar type of screen is the main communication device used on Krypton.  The first images of baby Kal-El appear on one of these: a screen housed in a floating oval construction made from metal that visualises data in sculpted three dimensional depth. Although it visualises the baby as it appears in the womb, the body of the foetus appears to pop out towards the viewer. These data interfaces are used at various points during the film and I think it is worth making a bold claim here: that they reflect the move towards 3D printing in recent years that is returning us to a sense of tactile engagement with data in a way that was overlooked with our focus on projected images since the birth of cinema and in later reincarnations such as television and computer screens. Sadly, the fact that I find this to be the most interesting part of the film suggests that Man of Steel will not become the more memorable of the Superman films.