This past Wednesday and Thursday at the Universal Hilton was the annual 3D Summit event. This event boasted some of the latest and greatest 3D talent and tools which were spread out amongst a myriad of key speakers and hallways filled with prototypes, proof of concepts, demonstrations and demos. Myself and Ian Vertovec were asked to accompany the Quantel booth and help demonstrate the new 3D options in Pablo v5.
I imagine for a number of people I know, a summit dedicated entirely and exclusively to 3D would be the last place they’d want to wander around a sea of blurry monitors and three-dimensiontal pitchmen. For others, the 3D box-office success that 2010 has shown made this a momentous year of justified 3D expansion and reinvestments-especially in the broadcast market, which was surprisingly in full array as compared to years past. I find it very interesting to research and observe the arguments about 3D from consumers, professionals, and the average filmgoer.
[Taking out the issue of profit for a moment] – I find there are so many different components going into making and exhibiting 3D pictures that I really don’t think basic conclusions can be made for or against the medium are as simple as “you like it or you don’t.” For example, as an artist, I find that animated films are absolutely superior in 3D and that the experience is personally much better than watching the same film in 2D. I have found this to be a fairly common feeling as substantiated by overwhelming positive consumer feedback Disney’s (soon-to-be highest grossing film ever) Toy Story 3D. As a digital intermediate supervisor, I find that 3D movies are a lot more complex, which makes them fun and challenging to coordinate and execute. Plus I enjoy the challenges of a 3D DI in that they offer some very unique opportunities for progressive colorists to show their skills as compositors and imagists as they prepare to master a film in nearly a dozen different configurations. But as a technologist, I think that 3D is all over the map, and the 2010 3D Summit taught me one thing above all: in 2010, the mark of acceptable quality for a 3D feature is not nearly as high as it needs to be. The culprit, I believe, rests within the controversial term; dimensionalization.
Dimensionalization is the process in which two-dimensional images (images shot with a single camera/single eye) are brought into a 3D space using space differential tools (such as rotosplines) in order to separated foreground elements from background elements to generate two distinct images from one source. Upon experiencing my first advanced conversion tests earlier last year, I was genuinely excited to experience the results. As 3D markets grow world-wide, the demand for 3D content must be equalized in order to stimulate further penetration (especially with broadcasters). But as 3D content is much more expensive, requires more expertise, and is more technically challenging to acquire, dimensionalization of 2D content is quickly becoming an Achilles heel in the 3D industry. For films that originated in 2D, I recognized dimensionalization as the only option a single-camera film has to be considered for a 3D market. But what bothers me and some of my colleagues is that now that people are actually planning on 3D distribution prior to shooting, they are still seriously considering dimensionalization instead of bona fide stereo capture. The result in our opinion is an inferior picture nearly 100% of the time as compared to stereo capture.
Here’s a seeminly obvious statement: I can only perceive what I can perceive.
The idea behind that statement is that without a side-by-side comparison, some people simply cannot see the difference between dimensionalization and stereo capture. The most important thing to understand in this whole blog is the reason for this is not because there is no difference. The reason is because people have yet to tune their senses through experience to identify the good 3D from the bad. I am confident this will change because the level of attraction to 3D images will increase with quality, not decrease. In other words, as time goes on, consumers will learn to identify high fidelity 3D from low fidelity 3D. Though they won’t be able to determine that dimensionalization is the reason for a poorly perceived image, they will be able to pick out the good from the bad. Evidence for this can be found in the similar roll-out of HDTVs. Over the past decade, consumers have improved their ability to determine the difference between standard definition pictures and high definition pictures and even the difference between 720p and 1080p. In less than 10 years, hundreds of millions learned the terms, learned what to ask for, and even learned how to see it, so will go their accelerated comprehension of 3D.
The reason this is a problem for 3D is that consumers are vulnerable right now to accepting 3D. Nearly every prime-time commercial break contains a feature film advertisement for a 3D release and even cable is starting to push their 3D channels such as ESPN and Discovery. Because they are genuinely interested and even excited, we must deliver them high fidelity. This is fundamentally why I am weary of all the 2D dimensionalization.
Consider this:Avatar comes out and pretty much breaks every film record that exists (more or less). But in terms of 3D fidelity, Avatar is basically a flawless representation of the best 3D ever experienced (at least the best I’ve ever experienced). So while all of us have experienced 3D images from James Cameron that are stunning, most of us have not seen anything that matches up. In fact, with many films capitalizing on Avatar’s well-groomed 3D road to profitability, dimensionalization is actually feeding consumers lower quality 3D, which will ultimately hurt the ability for 3D to stabilize. Contributing tho this is cheaper 3D televisions, automatic & realtime 3D conversion, and short-cut ways to capture 3D with cheaper 3D cameras is also lowering the 3D bar. Many manufacturers are cashing in, but some are at the expense of what should be an improving medium, not a declining one.
But it’s definitely not all bad: 3D Summit had some amazing technology and I experienced some major improvements to current 3D technology. From new 3D rigs like the Element Technica Neutron, to the Panasonic 3D 103” Plasma (breathtaking 3D display), to new and improved (and stylish) 3D glasses, there are some major developments that consumers and professionals are going to want to get ahold of. But peppered throughout the show floor were numerous dimensionalization examples – some good, some not so good, but all completely inferior to the real thing.
What I hope is that professionals consider that the consumer of 3D may not be as naive as they may seem. They may not know it now, nor can they put their finger on it, but the distinction from low and high fidelity 3D is going to become a key issue that will be a reason for the potential rejection of this transplant. We are obligated to delivery the best to the consumer – so make sure that tests for stereo capture vs. dimensionalization are done right and done extensively. This goes for the way in which the 3D mastering and color are performed as well. If you are of the camp that wants 3D to fail- just keep pushing dimensionalization and I am confident the magic will quickly wear off. If you want it to stay, then push the boundaries of true stereo capture. If you really put your own creativity to the limits, you will also see the difference and the difference, I believe, ultimately spells the acceptance or extinction of a three dimensional filmmaking and broadcast world.