A Genie for Uncle Guillon

He’s also developing his Préviz On Set system, with the goal of eventually catching up to the VFX On Set concept. According to him, this system is intended to be used more on the TV market than on the cinema market, at least in France, as a result of the need to work in studio and of the recurrence of effects). He’s also in the middle of working on an ultra-ultra-high-definition film (24K) that is going to be screened and integrated into the architecture of a major museum currently being built abroad.
Proof of how importance his work is and of his experience as a creator of images came in the form of the “Génie d’honneur” award that the Paris Images Digital Summit is preparing to award him, alongside Phil Tippett, the great legend of animated effects on Star Wars and Robocop. (FR)

How could you summarize all of the years you’ve spent doing special effects ?

Christian Guillon : We spent our time sawing off the branches we were sitting on. Films are always prototypes, and we’ve lived through thirty years of hybridization of a field in constant evolution. The result is that we’ve spent an enormous amount of energy tirelessly inventing procedures that end up being obsolete a few years later.
Another lesson is that, in the end, a major portion of what makes up our work is being progressively ingested by other departments (editing, colour grading, set design, camera…).

That’s a source of relief for us. Bit by bit, “invisible special effects” are becoming a part of the global production process. In my opinion, that confirms that special effects are consubstantial with cinema itself, and that cinema has always been a sort of special effect.

In your opinion, what deep changes has digital cinematography brought about ?

CG : Ever since digital became a part of every step of the production process, things have become considerably easier for us. No longer having to worry about getting images from one format to another is a great relief.
It’s a bit like mankind’s transition from Neanderthal to Homo Sapiens… The Neanderthal quietly evolved within its own limitations. Neanderthals were very skilled hunter-gatherers capable of building fires and carving flint, but who, despite their large brains, would certainly not have invented nuclear fission or mayonnaise. Enter Sapiens. They cohabit for a while. Were they able to mate and reproduce ?
For a long time, it was impossible to have digital and film-stock together : everyone insisted on only one or the other. Then came along the scanner, and hybridization began. We all have about 2% of Neanderthal DNA, but only Sapiens remained and brutally forced evolution to go up a notch. As far as digital is concerned, we call it a disruptive technology.
Sapiens hasn’t finished evolving, and isn’t the final human species. There are still things that remained to be invented in VFX, too.

All the same, it sometimes feels like there’s a bit of a digital-special-effects overdose, doesn’t it ?

CG : On one hand, as is the case with any technological advancement, one yields to the temptation of showing off one’s technical abilities and virtuosity : impossible camera movements, excessively phantasmagorical environments, plethoric armies, etc. In general, that tends to run its course.
On the other hand, people are seeking to imitate the old medium, by reproducing film’s artefacts (lines, lack of stability, flares, lack of depth-of-field, graininess and texture…) that are considered to be either a condition of (photo)realism or marks of culture. That, too, will pass with time.
Effects of depth-of-field have never been “realistic”. One might indulge oneself in nostalgia for anamorphic lenses (I personally loved them), or explore new vectors for the depth of the image, which must be developed within the visual grammar of the future.

But, in your opinion, what is it that characterizes this new digital age ?

CG : There has always been a debate between the image as a witness and the image as a composition. This conflict has always existed within the cinema community, but it is also present within each image.
Dissociating the elements that make up an image is as old as cinema itself. With the first special effects, overprints, additions of miniatures or mock-ups, it was immediately understood that an image can be composed by bringing together dissociated, artificially-created things.
Once again, sound was first, but with the greatest semantic discretion : re-recording never made André Bazin apoplectic (it did, though, Straub and Huillet, which does give me comfort).

Bit by bit, special effects accustomed us to dissociating the foreground from the background (green screen), the separate fabrication of painted backdrops (matte-paintings), and then calculated characters or objects (CGI hybridization), and has now come to the dissociation at the level of the actors (motion capture on a digital model of the character). Now, it’s even possible to record phenomena from the real world in a pure digital file format that can later be used as a reality vector by incorporating them into a calculated image. In the latest Star Wars episode, 80 to 85% of the image (in time/surface units) is made up of calculated images, so we’re really at a tipping point.

What film stands out in that sense to you ?

CG : Avatar is a technological and semantic landmark. It was the first time that synthetic characters were equal in status to filmed actors. They engage the viewers’ empathy just as much as the other characters do, thanks to the fact that they are animated by actors. I mean animated in the broadest sense of the word, but it’s the proof that calculated images, so long as they are enhanced by recordings from the real world, are capable of eliciting identification.
Given the state of technology at the time, James Cameron had the intelligence to make the CG characters different than humans, although they were humanoid, and ones that we had no prior reference to.
It makes me think of Jurassic Park, another ground-breaking film, in which Spielberg staged digital dinosaurs that no one had any reference to besides drawings or skeletons. At the time, a project based on CG dogs or tigers would have been impossible.
A few years later, The Life of Pi or The Jungle Book would demonstrate the incredible progress that synthetic images had made towards achieving photorealism. The next step will surely be a film in which the fusion between calculated and filmed human characters will have become undetectable. The bridge is being built over “Uncanny Valley,” and it will soon be opened to the public.

Are you interested in virtual reality ?

CG : In principle, this is still dissociation, but in this case, the dissociation is between the directing and the image-making. It will become possible to capture a real phenomenon, such as a scene from a play, in an exhaustive sense, and no longer simply from one point of view. I don’t mean 360°, which still involves a single point of view. Instead, I mean a type of image capture prefigured by technologies such as 4DView or Lythro.
That step will do away with the “director”, whose existence is inseparable from a single point of view, but will still require an “author” and a “metteur en scène” to write the screenplay, come up with the scene, choreograph it, and invent a scenography.
That exhaustive form of image-capture will become the material that one or several “directors” will be able to work with, and who will make their own linear version out of the scene by imposing their individual point of view, their editing, and their own camera movements (for example). But that raw material will also be able to be directly explored by the viewer, who will become a “spectauthor” or a “spectactor” and it will become a VR version. Some people refer to this “multi-revisionist” concept by the term “Deep Media”.
It makes me think of the work of Fernando Pereira Gomes, a Brazilian photographer who works inside the video game Grand Theft Auto V, within which he takes street photography. He shows his photos, but he has no camera (the digital-film quarrel is long behind us), and his “subjects” don’t exist either. He works just like Doisneau or Capa, he’s a photographer.

As far as I’m more modestly concerned, I recently realized that we were already working from within VR without necessarily expressing it that way. At “Les tontons truqueurs” we’ve created an immersive simulator in collaboration with Mikros’ R&D department, for the director of a film we were working on the VFX for. The film is intended to be projected in an immense and unconventional space, and I suggested that simulator to the director so that, whilst filming and editing, he might better be able to visualize how the images would be perceived by the audience in that particular room.
Today, we’re in postproduction on that project and we’re doing all of the VFX presentations with the immersion headset. In that particular case, how that tool might become relevant to us has become perfectly clear.
My conclusion is that VR is going to naturally become a part of our work in order to meet new needs and without becoming the subject of a theoretical debate.

(Interview conducted by François Reumont on behalf of the AFC, and translated from French by Alexander Baron-Raiffe)