I love the look of VIDEO
I LOVE the look of video, I love the aesthetic of the electronic image. I see chemical image of film and it just seems soft and dull and lifeless to me. I see the razor sharpness and the infinite flexibility of video, its density and dynamism and vibrancy and I think nothing but film is dead.
Now there's a statement to draw the ire of the purists and the technologically insecure. Should I go further...?
The film look is bullshit; a product of marketing representation and the digestible distillation of an association with a particular mode of viewing. The 'film look' is a cultural rather than aesthetic understanding; one drawn from our legacy of personal cinematic experiences in the movie theatre from a projected image. Thus, when it comes to making 'films' in the digital age for ourselves we innately want our films to evoke those same nostalgic memory associations we have with celluloid. This we translate as the aesthetic of film, the 'film look', but in truth it's more about cultural and personal association.
Certainly there is a 'film look', a set of visual characteristics derived from a medium made of celluloid and silver emulsion, but specifically seeking or choosing this 'look' because of a perception that it 'looks better' or is somehow visually superior is an argument very difficult to sustain on a technical or scientific level.
Film looks like Film, no doubt. The organic nature of its grain and distinct visual imperfections delivers a particular characteristic. But it would be a fundamental misnomer to then surmise that in the same vein Digital looks like Digital. Digital is not a medium in possession of innate characteristics as celluloid is. Digital is just Binary; representations, in Zeros and Ones, of visual information. Digital looks like whatever you want it to look like, so long as you know what youre doing and understand how to manipulate the Zeros and Ones.
So what really is the process of crafting the 'film look' in digital? It is an elusive thing. Strangely the 'film look' is often referred to as a specific technical element but that technical element is very difficult to qualify.
The simple truth is that any camera image source will, first and foremost, look like the quality of the camera lens. Shallow depth of field is often cited as key to the 'film look' but DOF is purely a product of lens and aperture exposure. Put a good fast lens with wide aperture on a digital camera and you can have every bit as much DOF control as a film camera. Defining the 'film look' by shallow DOF is technically bunk since DOF has nothing to do with recording medium.
The second element, much associated with the 'film look', is the cadence of its progressive movement; its visual rhythm. With digital and electronic video originally rooted in TV there was a long association with interlaced imagery. Interlaced images, by the nature of how they're assembled (as 50 or 60 fields rather than 25 or 30 frames) creates a distinctly smoother moving image lacking the slight staccato feel of film flicker.
Stu Maschwitz, one of the founders of
The Orphanage, developer of
Magic Bullet and author of
Pro Lost (one of the most information rich blogs on the net) wrote this about the relationship between Human beings, Flicker and Storytelling.
Video's frame rate being as close to reality as we can discern jibes with our ingrained perception of how video is traditionally used: to document real-life events. The TV news, reality TV shows, and our own home movies have a documentary quality to them that subconsciously suggests to the viewer that they are seeing actual events. Even sitcoms and soap operas are less like movies than they are like simulations of being in a studio audience watching a live performance. Video clues us in that we are watching reality, and by showing us everything, it invites us to passively absorb it. : OVERVIEW Movies are anything but reality. Ironically, by showing the audience less (40% of the temporal information of NTSC video), they trigger a part of our brains that works to fill in the missing information. In this way film creates a more participatory experience and at the same time informs its audience that what they are viewing is an authored, narrative work. This is backed up by our historical associations as well we have learned to associate film's flicker with storytelling and video's unflinching detail with reality.Since before history mankind has sat around campfires and told stories, and there are those who suggest that this association with narrative and the flickering image is so deeply ingrained in our collective unconscious that it in part explains our love for movies. Whether this is true or not, applying Magic Bullet to your video instantly transforms it from feeling like just another bit of DV camcorder footage to something more.But following Stu's argument, this association of the flickering image is a purely cultural one rather than one drawn on the basis of 'quality' which is so often cited with film. Similarly, and by contrast, the 'undesirable' smoothness of video is not a product of digital/electronic means itself but rather of the Electrical Power and Broadcasting infrastructure that traditionally supported it. US NTSC uses 60 fields per second for no other reason than US power grids use 60hz oscillations. Most of the rest of the world on PAL uses 50 fields because the power grids are at 50hz oscillations.
Again, digital images, of themselves, have no innate 'look', they are just data manipulated and its the manipulation that defines a 'look'. As the old infrastructures, that have traditionally defined much of the aesthetic, dissipate the manipulation of digital data is left more to the filmmaker than the infrastructure of electricity and broadcasting. Hence we have the dramatic shift in recent years to Progressive Scan cameras, 24p, 25p and 30p shooting as well as the very 'filmic' technique of over and under cranking of frame rates.
From there the rest of what defines a 'film look' to the common observer is the colour and tone of the image; how the media is treated, processed and manipulated in post-production.
In the long history of celluloid production this processing of colour and tone was a photo chemical process; a manipulation of the visual information by means of chemistry. But in truth such processes where not particularly common with the tone of an image being largely pre-defined by type of film stock selected and the manipulations of exposure in-camera during shooting.
Whilst there is indeed a long history of such manipulations in cinema, the process of colour grading (or colour timing) as a common, widespread and accepted part of post production is a relatively recent development - a shift that has seen the practice move from the fringes of experimentation and special effects to simple mainstream commonality.
Arguably one of the preeminent focuses of colour grading processes in the digital age has been on getting digital to emulate the visual characteristics of celluloid. Such choices impose on digital cinema two categories of manipulations; the recognisable characteristics of film which are otherwise absent (namely grain and flicker), and particular colour tonnings for style and mood (tone, colour wash, contrast and so on). All these are focused in popular perception on the 'film look' and 'better' visual quality.
Yet there are distinct conundrums and contradictions here. The artificial insertion of celluloid artefacts of grain, organic emulsion, removal of frames to force 24p, jitter and flicker are all acts of deliberate degradation. Any way you slice it, putting such elements into an image where they don't previously exist is an act of degrading and lowering the visual quality of the image. An overtly strange act when the intention is to get the 'film look' because it looks 'better'...
The second set of process actions associated with colour and tone are those often designed to emulate the particular chromatic properties of specific film stocks. But this is really just a process of using a film stock as a reference point. The digital colour manipulations in grading processes far exceed what is possible from film stock itself. Digital image data is simply that, data. It's an almost infinitely flexible set of data waiting to be given a 'look'; a concept fundamentally divorced from film stock which has an inherent 'look' based on brand, type and chemical make up.
At the risk of public lynching from the film purists I'd argue that in the digital age, film as a medium is a distinctly anti-creative format. By its very nature shooting on film limits, restricts or cuts off the filmmaker from a host options creative options that would otherwise be open to them. Celluloid is not a blank canvas, not an open slate onto which to paint with all the available colours. Digital, by its technical make up, a blank and unformed ball of clay that can be shaped into any conceivable form.
I cannot help but be confronted by the irony that as we, creative cinema makers, are handed the most flexible form we have ever know, one unrestricted and infinitely open, our first overriding instinct is to degrade it, limit it, deform it to enforce upon it the restrictions of its predecessor.
Perhaps this is simply a techno-aesthetic derivation of Bolter and Grusin's theories of 'Remediation' where by new media begin life by replicating the tenets of old media before finally breaking free to find unique properties. Photography remediated painting until it found its unique paths; Cinema remediated Theatre until it forged new languages; it seems digital media as a production format is destined to remediate celluloid media until filmmakers embrace and/or discover the unique properties and possibilities of digital as its own platform - one that has shaken off the shackles of celluloid limitations.
Until then a Google search will continue to reveal the term 'film look' as one of the most common discussion topics amongst filmmakers.
I may well be the only one but I can't wait for the day we all 'get over it' and stop seeking to limit and curtail the evolution of the moving image and focus on exploiting its new properties. Colour, style, form, look, visual delight are what we should be aiming for, . the 'Film Look' for it's own sake is bullshit.
Posted at 12:00AM Mar 17, 2008
by Mike Jones in video |
Posted by editblog on March 17, 2008 at 01:41 AM EST #
Posted by ouin-ouin on March 17, 2008 at 03:43 AM EST #
Grain is part of the appeal of film, but simulated grain is a very different thing. It is essentially noise, whereas film grain is signal, same as pixels but randomly distributed. And because of the random distribution, each frame of exposed film will hold information unique to that frame, adding to the perceived resolution of the moving image.
To emulate that in video, you would need to start with an image many times the resolution of film. And if you have such an image, it seems a bit silly to degrade it.
This is also why I detest the overzelous use of noise reduction on DVDs — it removes detail and leaves a sickening blur behind. Film grain is not noise. It can't be emulated by adding noise, and conversely it shouldn't be filtered out with noise reduction.
I agree that digital images have the potential to be so much more than a poor man's film, but I for one would like the two to coexist.
Posted by EditDroid on March 17, 2008 at 07:12 AM EST #
Congratulations, and I hope you can profit handsomely by your skills as both a writer and a producer of creative images.
Posted by John Meyer on March 18, 2008 at 05:23 AM EST #
Posted by Mike Jones on March 18, 2008 at 10:20 AM EST #
Your assertion that digital images have no innate look in themselves is not true however. Even the RAW data from a digital camera is the product of it's sensor. And different manufactures make different sensors. And they all have subtle differences. The idea that that a digitally encoded image is *untainted* is simplistic. They all have differing and measurable signal to noise levels, that will affect the way the image is encoded. Colour science also differs from each manufacturer. So even RAW images will be take on the inherent characteristics of the sensor with regards to noise, colour reproduction and dynamic range. So although it is just DATA, each camera produces a different set of DATA and a different look. Just like film stock.
Your misleading history of colour timing is simply not true. It has been possible to colour grade films for many many years. In fact the whole use of the phrase *timing* comes from the optical process that colour grading was and is still done by today. And for that matter, so are the terms one light and best light. Day for night looks, not to mention scene matching and sheer creative grading is very easily achieved and created using this process. This year's AFI award winning Romulous my father was all finished photochemically aside from a single 4K VFX shot. Obviously it can't be all that bad.
What digital processing has made possible, was much finer control of the colour correction process. And this perhaps is what you meant when you referred to as, "more recently". Pleasantville (1998) was probably one of the first films to be graded in this manner, but mainly because every single shot was a VFX shot by nature of it's storyline. Oh Brother Where art Thou (2000) is considered to be one of the first DI or digital intermediate films graded in this manner, where colour effects were achieved that weren't previously possible using photochemical grading. Both films of course, where shot on film.
And this is my next point.
How is an image scanned from a film negative any less *digital* than a digitally acquired image ?
Once the film frame is scanned it is JUST as digital as a digital image rendered by a digital camera. It starts from exactly the same point as a digitally acquired image. If you argue that film is anti- creative, then all of your arguments about it's post production workflows are neutered when you consider that a scan of a film frame can have all the same image manipulations as a digitally acquired one. A flat LOG scan of a film frame is just as flat and dull as a RAW image from a still camera. The same blank canvas. So let's assume then, that a scanned film frame can be just as digital as a digital acquired digital frame.
What's anti-creative about film in production ? Cost. ? Setup time ? availability ? These are all valid arguments to some degree. But it's also not as simple to look at a single cost on the film's budget and point to it and say AH HA !
While film stock and processing is expensive compared to acquiring digitally, if you factor in the cost for the higher end digital camera platforms, you'll find that the post production infrastructure is significant.
The much championed RED camera, which uses the D-SLR approach to imaging provides some very fine useable images with it's own unique look. And then there's the terrabytes of data it generates. Plus now you have to back all this data up in a safe and redundant way, because you can't go back to the negative for a re-scan if a drive fails. And then there's the actual infrastructure of trying to even edit and view these files. To do it well and safe costs. And again, I would say that it's inherent look is great if that suits your project, and it's as digital as a digital scan of a film frame so all the post grading and VFX work well.
On set it's reduced dynamic range compared with film, like with all digital cameras, means that you in fact often spend more time lighting because you need to ensure that you get it right on the day. Film allows us to be somewhat lazy in this regard. Now this is not film look. This is being cinematic.
The idea that digital is faster to light and requires less light is also a fallacy. Sure, you can turn the camera on and you'll get a picture right away without any lighting. And the same goes for any film camera really. If you want it to look *cinematic* (not filmic) then it still requires a great deal of care and attention on set. You still need a dolly or a steadicam. You still need to be able to light for the camera. You need the same crew and level of experience.
Plenty of people are happy to use digital for what it's great at. Russian ark (2002) for it continuous long take. Films like Blacktown (2005) couldn't be made without digital shooting technology allowing intimate access to non professional actors.
Progressive displays have been around since..well computers, and we've been using them for at least the past 20 years. But most US Drama has been film acquired over the years. We tend to associate video with lo-fi and disposable, simply because that's what most news footage is and that's how we consume it. And it's natural to associate that with *real*.
By dissing the film look, you're actually engaging in the same debate you're pretending you don't care about. The digital mantra that digital is better leaves behind a lot of valuable film-making techniques that lots of people don't seem to be in a hurry to leave behind even though digital film origination as been with us at least a decade.
I think it's naive to dismiss any form of image capture, and hitch your wagon to any one acquisition platform, they all have a place. I recently shot some sections of a film finished to 35mm on a mobile phone.
If this revolution is happening and has been happening for the last few years, why hasn't there been a bigger rush ? Why does all of Soderburg's film originated films sell squillions of tickets while people don't even know he's cranking out other digital films....
How about we just use the best paintbrush we can skilfully use and afford to paint the best picture we can for the story we're trying to convey.
jb - I love the look of what ever the story teller has chosen to use....
*i wait with bated breath to see what David (i'll never shoot film again) Lynch will do after his underwhelming PD150 shot Inland Empire.
Posted by John Brawley on March 18, 2008 at 07:26 PM EST #
To me the creative strength of the moving image its its flexibility, its malleability. The equation I make is simply that the More flexible the image is, the more creative it becomes. The more it can be shaped to a desired impetuous and outcome, the more innately artistic the medium. This is where you are dead right to refer to DI process and delivering unto celluloid the same flexibility of the native digital. You raise superb and valid points throughout and I most certainly appreciate the thought and effort you put into the response.
Posted by Mike Jones on March 18, 2008 at 08:17 PM EST #