Blue Sky - The Cloud Hosting Company

A rationale for CGI luddites


Why even good CGI can annoy...

CGI luddites

Before I go any further, I should put the title of this piece into context by saying that I am NOT a CGI luddite myself. I have been doing 3D modelling and illustration for nearly 15 years in C4D, Max, the continually-amazing ZBrush and many other apps, and can honestly count the emergence of the T-Rex in Jurassic Park as a life-changing moment. So what follows (in case you can't be bothered to read it but find the title irksome) is not a luddite rant against voxels, just a few thoughts about the way we perceive visual effects, and why CGI continues to bother so many people in comparison to previous methods of screen magic...

It's hard for the senior geek to comprehend the expectations and conception that younger viewers - Hollywood's target audience - have about visual effects in movies. We who were old enough to be familiar with movie culture in the years prior to the ground-shaking advents of T2 and Jurassic Park may have had the excitement of watching a rare sea-change in the future of movies unfold before our eyes, but perhaps we can't get the full benefit of having been born into the post-photochemical world. We may have a curmudgeonly and nostalgia-driven affection for the 'old ways'.

If it was already hard to find a common frame of reference for discussion of visual effects, it's getting harder, since the terms 'CGI' and 'visual effects' are becoming lazily interchangeable; even on supposedly knowledgeable DVD commentaries I have heard people refer to pre-digital visual effects (i.e. in 1970s, 1980s movies) as 'CGI'. Perhaps even lazier is the far more prevalent alternative moniker 'digital effects' (technically film has been potentially 'digital' since the emergence of the AVID editing system in the early 1990s, and the term 'digital effect' can as easily describe the compositing of two 'live action' elements as it can a rendered 3D mesh).

But if there is still controversy, at least some of the confusion that led to the great Paul Verhoeven's typically spirited commentary on Starship Troopers (1997) has abated...

The 1990s were the frontier 'wonder years' of CGI, before zBrush, real-time organic modelling and Moore's Law ushered in the age of forest elves, closing off perhaps the most prolific decade for 'straight' science-fiction films since the heyday of atomic-age movies. It was in that very experimental decade that CGI earned its most vociferous detractors; not only had 1993's Jurassic Park brought textured rendering to the movies, but had done it with such a dazzling show of vanguard action that even the imitators that followed over the next seven years (who had better software and render-power) were trumped by smaller projects such as Toy Story, which understood what the 'state of the art' at that time really was for computer-generated imagery.

Twister (1996)We were long past the dreadful blocky figures of Dire Straits' Money For Nothing video, but still a fair way from the Virtual Reality scenarios dominating the films of the decade, such as The Matrix, Johnny Mnemonic, Existenz and Total Recall. Effects units on many of the day's blockbusters were glad to have at their disposal tools that could render what the miniature and photochemical world had always struggled to, such as water (Deep Impact, The Perfect Storm) and weather (the absurdly successful overgrown TV movie that was Twister), but remained surprisingly faithful to miniatures, even for the kind of work - such as starships flying through space - at which one might have presumed CGI was already adequately convincing.

There are various reasons for that: Hollywood's suspicion of the unproven; the considerably fewer quality effects houses of the day, all charging a fair amount for work that was more prized in its nature and novelty than necessarily its quality; or, as Jean Pierre Jeunet ultimately decided for 1997's Alien Resurrection before engaging Matthew Gratzner and Ian Hunter to create the models for that movie, the not uncommon view that most CGI still looked pretty fake, even when depicting the bizarre - which should be its obvious forte.

Lord Of The Rings (2001)Eight years after Jurassic Park, Lord Of The Rings was perhaps the first movie to really advance the field of CGI; its aesthetic needs were rustic and organic photorealism, not the clean edges of starships or the Bryce3D-style CGI sunsets that made the 'golden hour' last all day in Hollywood's effects houses. Peter Jackson's take on Tolkien began to succeed where Jar Jar Binks had failed in 1999, and the age of character-driven CGI was to sweep all before it. John Dykstra's CGI Spider-Man in 2002, following up the popularity of 2000's X-Men, brought the superhero genre into what proved, and continues to be, a two-horse race at the Hollywood box-office.

'Photo-real' was the lodestar of visual effects since the days of Meliés, not just a recent objective for CGI. Directors such as John Carpenter declared that the technology was therefore 'just another tool', and that if it was fake, so then were the prosthetic dummies and kitbashed models of old. In the 2000s, 'photoreal' became a frequent enough occurence, most especially in summer tentpole disaster movies, as to potentially draw some kind of line on the 100 year race to achieve it.

Why are there are two disgruntled directors, young and old alike, for every John Carpenter? I listen to enough disc commentaries to frequently hear the use of CGI disparaged in professional eschelons. And if Tarantino's CGI moratorium on Death Proof was some kind of a retro stunt for the 70s-obsessed director, he's certainly not alone in the thought that inspired it.

And much as I love to use 3D software myself, much as I thrill to the best use of CGI, I too feel that something is dying in movies because of CGI, and I just want to think about what that might be - if it's anything but a nostalgic whim on my part.

'Viewer pessimism'

Will Smith shoots 'Hancock' (2008)One factor is what I can only describe as Viewer Pessimism: the assumption that anything that is difficult or which would be highly effortful or faintly risky in a movie has been mimicked with CGI, even when it could have potentially been done 'real world'. Even when it was done in the real world.

My own most startling experience of this was in watching some of the before and after versions of effects shots for 2008's Hancock; in one particular shot, Will Smith is in the last phase of leaping from the sky to come to the aid of police dealing with a hostage situation, and since he is 30 or 40 feet off the ground and zooms in for a smooth landing, I figured he was only 'real' from about two foot off the ground. In fact the 'raw' footage reveals that Smith was doing some pretty impressive and acrobatic rig-work in a shot whose only concession to CGI was a little wire-removal. I'm guessing that most viewers would have read that shot as at least 50% CGI, and it's even possible that director Peter Berg, consulting with VFX wizard John Dykstra, went 'practical' in this case for reasons which had nothing to do with CGI's ability to make the shot happen.

"The background ambience of CGI in movie culture risks to put us into 'cartoon' mode even when there's not a shred of CGI on the screen"

Viewer Pessimism then is when you mistake something real for a CGI 'trick'. And here, perhaps, is where a generation born and steeped in CGI may arguably get less enjoyment than the elder geek from the knowledge of watching an extraordinary 'real world' event - because they may not be able to identify it, if they are vague on the chronology of CGI or film history.

This spectacular stunt from 1975's Bond outing The Man With The Golden Gun is probably not something any director would put in a movie these days, because for CGI it's not 'wild' enough, and Viewer Pessimism might cancel the value of doing it for real...

And what sane director these days would endure the 88 sky-dives that Lewis Gilbert filmed in order to get the spectacular conclusion of the pre-credits sequence to 1979's Moonraker...?

Again, no-one would 'believe' it anyway, were it to be done now. The camera has never lied the way it can now lie. So Viewer Pessimism potentially robs us of the thrill of spectacle, of what's 'theatrical' about movies; the background ambience of CGI in movie culture risks to put us into 'cartoon' mode even when there's not a shred of CGI on the screen.

Death of the 'VFX Threshold'

There's one type of visual effect which remains unchanged by the advent of CGI: the effect that no-one notices. By definition it is usually relatively mundane in nature, and, much as the VFX wizards have always claimed to feel 'rewarded' by viewers not recognising their contribution, these really are the least-appreciated of all effects shots.

For instance, I didn't know until a recent listen to Nick Meyer's DVD commentary that the entire front and top section of this real-life set from Star Trek II: The Wrath Of Khan is in fact a species of 'hanging miniature' sitting very close to the camera lens - pretty much the oldest trick in the special effects world...

Star Trek II: The Wrath Of Khan (1982)

Because one presumes that a Star Trek movie has some kind of reasonable budget for set-building, and because the foreground model does not hugely widen the scope of the set, stagger any budgetary expectations or defy any natural laws, I was fooled into thinking that $15000 was spent on set instead of (perhaps) $4000 on a miniature.

Likewise CGI compositing and clean-up was used in Wolfgang Petersen's The Perfect Storm to raise the tide level of the Andrea Gail departing from the harbour. Here VFX merely created something that could have been shot for real if the shooting schedule hadn't gone awry...

The Andrea Gail departs in 'The Perfect Storm' (2000)

For such shots, whether photochemical, miniature or CGI in nature, no-one can care, object or appreciate - because usually no-one knows. These shots are outside our 'VFX Threshold' - the crossover point at which we are willing to 'Get out and push' in order to help the filmmakers tell their story, if we are enjoying the movie, or at least want to redeem the price of our ticket.

It's our ability to distinguish between what is possible to film 'real' and what must necessarily be 'faked' that has been eroded by CGI. In theory, forcing a greater suspension of disbelief in the viewer is an excellent development in the cinema-going experience. In practice it not only forces 'cartoon mode' in the viewer for what is 'real', as noted in the previous point, but also makes it very difficult for us to appreciate 'good special effects' - VFX sequences which could clearly not have been filmed 'real' (spaceships in flight, destruction of cities, etc).

"Just as the wildest stories refer back to the mundanest truths and experiences of our lives, so do visual effects need some grounding in our experience"

Avatar (2009)The moviegoing experience is not as 'passive' or lacking in imagination as some critics have suggested. We arrive at a movie with at least an initial disposition to do some work for the sake of our own enjoyment, to collaborate with the film-makers. We bring to a movie our sense of awe (or what remains of it in our dotage) and our own imagination. But just as no-one can bench-press for a straight two hours, the erosion of the VFX Threshold may lead to fatigue in our ability to suspend disbelief. At the end of the day, disbelief is quite a weight to sustain when (in CGI's more extravagant outings) all 'real-world' references have been removed for sustained periods.

For instance, with the exception of NASA footage, we have no idea of what a space environment looks like, and since NASA's output is so particular and the technology depicted in it so iconic, we are for all intents and purposes (Apollo 13 aside) devoid of any meaningful reference material with which to decide if the 'space movie' that we're watching is 'realistic' or not. Hence 'sound in space', and the lorry-like rumbling of large vehicles tacked on to space-behemoths to form in us some kind of association with a 'real life equivalent' of the fantasy we are watching.

Just as the wildest stories refer back to the mundanest truths and experiences of our lives, so do visual effects need some grounding in our experience. Our real-world experience is the hook to firing our imagination, the springboard from which awe departs. Without the VFX Theshold, we never get to 'relax' between scenes, and the CGI-driven movie can become a reductionist sandwich - without any bread.

Of course, none of this is an obstacle to the enjoyment of a good story well-told. But then, that's the very rationale by which one can also defend the appreciation of a good book - the most imaginatively collaborative of all story-telling methods.

Tactile empathy

The Greg Jein mothership from Close Encounters Of The Third Kind (1978) on display in the Smithsonian InstituteOne way or another we all play with models as kids, from the rubber duck onwards to dolls, play houses and action figures - and this may explain in some way a sentimentality for miniature work in movies as opposed to CGI models. A 3D mesh is teleological in nature; unless it gets 3D-printed, it's only on paper, and virtual paper at that. The aim of teleology is conception, templating for real-world things, but CGI never leaves the teleological realm in any professional sense. There is nothing to play with; there will be no memorabilia to bequeath or put in a museum; even the out-of-date diskettes containing the obsolete 3D format models of CGI classics are only real-world signposts to non-world, virtual assemblies. Even the word 'virtual' has a pejorative sense: 'Not quite there', both literally and - perhaps - figuratively.

One could argue that it takes an extraordinary imagination to engage meaningfully with the unreal and the genuinely abstract, a capacity above average - a child-like sense of wonder, which remains pretty much exclusive to children. Here I return to my earliest point - that those born into the CGI age, even if they tar all visual effects with the CGI brush, are perhaps best-fitted to appreciate them, since their earliest movie-going experiences were shaped by CGI.

Robophobia and the double-edged sword of CGI mystique

Anyone who has ever been replaced by a computer, or by someone better able to operate a computer or a particular software, has participated in the classic sci-fi scenario of cybernetically-driven obsolescence. Without an understanding of how a 3D pipeline operates, and how intrinsically uncreative CGI software is in itself (which trait it shares with all software, though CGI software is amongst the most impenetrable and unintuitive of them all to the rank beginner), there remains on CGI a taint of 'automation'. "What was once done by men is now done by computers!".

This at least is totally fallacious, since the men and women who currently participate in a CGI pipeline are analogous to the astronauts who emerged from the ranks of test pilots: not only do they require the same creative skill as their antecedents, but also an advanced practical and mathematical skill-set. Directors and visual effects artists often attempt to reiterate this in interviews and commentaries, but usually they're preaching to a very small and already-converted audience.

The facility of CGI software to accomodate the creative urge is a factor here. Not every gifted artist is a natural technician, and in the early days of the craft, it's quite possible that left-brain-centred operatives -  those with a more abstract interest in number-crunching than an urge to conceptualise and create - were among the most valuable proponents of CGI. If CGI's public mission to fulfil on-screen photorealism is at an advanced stage, then its mission to make its tools intuitive and accessible still lags considerably behind.

Pixologic's zBrushEven the most non-mathematical, intuitive and perhaps the most brightly-starred 3D modelling software, Pixologic's zBrush (used in the Lord Of The Rings trilogy, Avatar and practically every motion picture with a character-based CG pipeline since the early 2000s) has such a steep learning curve that the makers provide an extensive array of introductory videos and tutorials to get the curious past the initial pain barrier.

It's too simplistic to declare that CGI workflow is currently dominated by number-geeks when it should be dominated by 'creative' geeks - and that would be a terrible disservice to the legions of dedicated VFX workers who have carried and nurtured their creativity through a long and previously-unnecessary technical indoctrination to CGI. But the very same opaque mystique which awes the fledgeling CGI worker is what convinces many viewers that CGI is more 'science' than 'art'.

Explaining the CGI workflow in a simple and meaningful way is currently still problematic. Whereas it's easy to say "We built a model of The Poseidon and tipped it over in a big tank. Oh and we filmed it in slow motion to get the water scale right", it's rather harder to understand the extrapolation of vector-based curves into polygon-based meshes, the skinning, rigging and retopologising of geometry, the conversion of mathematical shaders and reference bitmaps into normal maps and displacement maps, the trade-off between the self-bouncing rays of ambient occlusion vs. prohibitive render times vs. network render resources...

"We did it with computers!"

Hollywood loves to dazzle, so it's no surprise if documentaries and DVD extras aren't afraid to make CGI look difficult. It is difficult, even when you know how. But all the viewer-respect and perceived value-for-money obtained by glorying in this mystique is perhaps counter-balanced by how 'alien' and notionally antiseptic a CGI workflow is. What inspires awe can also be hard to love, as any stereotypical Victorian child can tell you.

No more heroes...?

Phil Tippett, who evolved from stop-motion to CGIThe age of the putative 'auteur' or iconic VFX artist seems to be over. Many of those who made their names in the 'iconic' days of VFX - such as John Dykstra, Richard Yuricich and Phil Tippett - adapted to the digital era and continue to enjoy the pioneer-factor of their names. But it doesn't seem that their likes will emerge again in a trade that is now so subdivided among specialists that a VFX supervisor's role is arguably less 'hands on' than in the days of photochemical visual effects.

In fact most of them would argue that their leading role was always that of project managers, and the best willingly share their credit with the huge teams behind their greatest 'classic' successes.

But if we want any more stories about how a movie wizard performed movie magic from a garage, we are unlikely to get it from a VFX supervisor anymore, esconsced as most are in air-conditioned office suites, presiding over 50+ specialist CGI workers. Such figures can still steer and guide visual effects shots, but can't usually directly intervene mid-flow. Most of the serendipity involved in CGI VFX arguably needs to take place at the conceptual stage - most accidents any further down the pipeline tend no longer to be 'happy' ones, and one can imagine the possibility of a sterilising and cautious regime under such circumstances, with deadlines looming and budgets tight.

Gareth Edwards creating 'Monsters' in his bathroomThe lack of new VFX 'garage heroes' perhaps explains the huge interest in Gareth Edwards, a British VFX supervisor and technician whose debut directing feature Monsters (2010) not only showed extraordinary imagination and poetry on an absurdly minimal budget, but who also undertook the entire post-production CGI process himself using off-the-shelf retail software. It's a truly 'retro' experience to check out the 'Monsters VFX' documentary on the DVD release and watch the man who is going to direct the Godzilla reboot playing with a toy-plane and a video camera in his bathroom whilst trying to find the right look for an underwater lighting effect.

Needless to say, Edwards' bathroom is likely to be restricted to the usual activities during Godzilla. But just for a moment we saw an auteur 'doing it himself', 1970s-style. There's a personality, in every sense, to Monsters that can't possibly be recreated by 50 people in a CGI workflow, and perhaps it's only VFX technicians who turn successfully to directing who can give us that sense of a movie's 'unique' vision. And, if they're successful, they can probably only do it once, unless they're addicted to a bohemian lifestyle...

See also:

The last great frontier of CGI: The eye
Peter Jackson easing 3D eyestrain for The Hobbit?
Star Trek Remastered and the subtle art of CGI
The dismal future of 'R'-rated fantasy and sci-fi movies
The Praxis effect: Star Wars > Star Trek
Hollywood's determination to sell us Victorian 3D technology
The worrying rebirth of 3D
How to avoid getting a 3D headache in 3D movies
Six unlikely changes for the Blu-ray release of Star Wars
The iron-fisted father of movie matte painting
Secret immunity to 3D movies
Exclusive interview: The worlds of Roger Christian
Exclusive interview with VFX wizard Brian Johnson
Exclusive: 'UFO' director talks 'miniature revival' on Shutter Island
Exclusive: Lost retro spaceships from the 'Dan Dare' TV project


If you're interested in writing for Shadowlocked (disc and screening reviews, etc, or just getting some extra coverage for your extraordinary writing talent, get in touch with us.



#1 Eye Gouging Craig 2011-04-08 09:38
I love great CGI work but I really hate Poor CGI work.

If an object doesn't have proper motion, doesn't obey the physics of motion, its noticable and can really pulls you out of the experience.

Mouths are the worst of the lot. See Tron:Legacy and the digital Flynn. Whenever he speaks it just doesn't look right. The lips don't move properly. In "I am Legend" why do the creatures mouths open so wide, and whats up with the jiggle the jaws have. It is distracting.

In the original spiderman, a cgi fight scene is obviously cgi when gravity affects him at different rates in the same scene and his motion just becomes off and you say to yourself, thats cgi.

Good ones like Jurussic Park never had me say "Thats CGI" it seemed seamless.
Avatar was another, Gollum in LOTR another.

If your going to have real world object depicted by CGI, make sure the motion is right and we will never know the difference, nor would we care.
#2 RE: A rationale for CGI luddites Chris French 2011-04-08 21:59
One word: _JAWS_. Would that movie have been nearly as successful if Bruce had been visible as much as Senor Spielbergo wanted? (Hint: _Deep Blue Sea_....)

When any visual effect is easily made, visuals cease to have meaning; and we are forced back to Plot, Character, and Acting -- one *cannot* CGI Quint's "_Indianapolis_ monologue".
#3 Retcon much? Duder NME 2011-04-09 19:40
Once again, LOTR is propped as a piece of import at the behest of Star Wars, whose motion capture/digital characters set the standard years before Jackson entered the field, yet only because of their unwarranted lambasting by so-called fans does the also-ran LOTR gain any noteriety due to simply not being the former in the myopic bias of the genre fanbase. Hilariously sickening. It's only been... what... a decade, internet? Keep up the rewriting of history.
#4 RE: A rationale for CGI luddites erlend 2011-04-10 12:47
A major problem is motion - CGI artists seem to love unnaturally smooth motion. That's why I love the way Guillermo Del Toro combines practical and CGI creature effects - animatronics and prosthetics are so rigid in their movement and CGI is so smooth, that when you combine them you're not really sure what you're seeing.
I grew up with CGI, but even as a five or six year old child I remember being freaked out by "digital doubles," weird textures, glossy rendering and overly smooth motion.
Motion capture (or performance capture or emotion capture... all that shit) is the worst. Actors tend to overact with it, and in the final output their mouths never look natural and their eyes seem "dead." Even in avatar, i sometimes had this problem.
#5 RE: A rationale for CGI luddites erlend 2011-04-10 12:47
A major problem is motion - CGI artists seem to love unnaturally smooth motion. That's why I love the way Guillermo Del Toro combines practical and CGI creature effects - animatronics and prosthetics are so rigid in their movement and CGI is so smooth, that when you combine them you're not really sure what you're seeing.
I grew up with CGI, but even as a five or six year old child I remember being freaked out by "digital doubles," weird textures, glossy rendering and overly smooth motion.
Motion capture (or performance capture or emotion capture... all that shit) is the worst. Actors tend to overact with it, and in the final output their mouths never look natural and their eyes seem "dead." Even in avatar, i sometimes had this problem.
#6 RE: A rational for CGI luddites Dinoslay 2013-12-30 13:09
Before computers came around the creation of visual effects was prohibited by budget and physical limitations. While the creation of VFX these days remains as difficult and demanding of artistic vision as always I think the supporters of both practical and CGI miss something in this debate.

Audiovisual entertainment is in a sense defined by visual gimmickry. You could almost compare it to a circus act that different audiences pay to see. The point of the format is to entertain, to distract people from everyday life. What's missing, however, is a certain type of reciprocity between circus and theatre. IMHO, the best film experiences of the past and present alike have a harmony between story of well written character(s) and visual sophistication.

To repeat myself, the creation of visual effects is in no way less demanding today as it was back when all that people had was different materials and good ideas in how to use them for the best visual result that could be produced. The only difference is in how the old prohibitions have become redundant due to computers taking over certain functions. To most the loss of such limits sounds like a good thing.

Truth is, that's a double edged sword. Some people make cheap gimmicks, others do expensive ones. Difference of quality aside, it is nevertheless gimmickry. The less prohibited creation of visual gimmicks has become a foundation of sorts to film-making, the price being that storytelling suffers from being outweighed by said gimmickry. It doesn't really matter what type of stories you try to tell when the end result is always the same; the gimmickry wins. Is that the sole motivation to build a story around? I don't think so.

I know that this sounds harsh and I by no means want to jump the gun and offend the passionate but it has to be said; film-makers, watch out! If you're not careful with your beloved format a bigger moronic need will lead to a bigger moronic supply. As a result, bye bye storytelling, hello attention deficiency.

Report an error in this article
Add comment (comments from logged in users are published immediately, other comments await moderator approval)

Shadowlocked FULL TEXT article RSS Shadowlocked RSS