Exactly, i need to finish my film though, I'm feeling 99% BS not having finished a feature. I think to get a significant draw people want a spectacle and have a blast.
It's for sure faster shooting with convergence in post if you know what your doing in production and are going to keep that convergence (also is nice for depth of field). Problem is, the convergence has to get tweeked a lot in post, thus the need for some of the crazy functions in nukex/ocula. But ... I think your right, and the fact that a lot of people are doing the 3D completely in post really is saying that it's often best not to set the convergence during the shot. (obviously all the vfx convergence is done totally in post). I would love to do the setup like you stated, would speed up my workflow a lot. BUT what I spent all day today/Sunday doing, was taking the shots for a scene with one camera, then making sure my lidar of the scene was good so I could speed up my 3d process in post.Thanks for the warm welcome Patrick - and the encouragement to "comment away."
Geronimo!
For starters, I have a very strong conviction that many of today's standard practices for shooting live action 3D should be abandoned altogether. I know this will be found controversial, but jumping in with both feet, I'll continue by saying that, in my opinion, both selective focus and convergence are counterproductive to recreating a perception of reality, as compared to how we experience stereopsis naturally. These practices won't easily be put to rest, even with 192 fps and 16K.
First, I'm suggesting that instead of fixing the problems caused by use of convergence - with floating windows, CGI backgrounds, vignetting, selective focus, and any other tricks that can be applied in post to hide the byproducts of convergence - just shoot with zero convergence. Keep the lens axes parallel at all times.
I'm most likely just revealing my ignorance of factors that make convergence "necessary," but I can't help but think its use stems from people thinking we should emulate the convergence applied by our own eyes in the real world. I contend that what our eyes do (converging for subjects that are closer than 200 feet - yes, they come out of parallel even for subjects that reside that far away) and how we actually perceive the world when our eyes are converged are two different things.
Hold the tip of your index finger about 4 inches in front of your face. Focus on the tip of your finger. While maintaining that point of focus, can you say that the doubled-up background has ever disturbed you in any way? The answer is "No" because you can keep your finger right there in front of your face and readily de-converge your eyes to focus on something behind your finger at will. With natural stereopsis, nothing forces you to converge your eyes to only one plane along the Z-axis.
You and I and a third person can stand along side each other in a forest, with a tree trunk at a distance of say 10 feet, partially blocking our view of the trees that lie beyond, with a small butterfly bouncing along in the air about three feet in front of us. Nothing, absolutely nothing, prevents me from focusing on the most distant trees in the forest, while another person simultaneously focuses on the bark of the closest tree, while you simultaneously focus on the passing butterfly. This is how we experience real life - rapidly changing the convergence of our eyes to explore different planes along the z-axis at will. And no one is disturbed in any way by the doubling up of background subjects that naturally occur with convergence of their eyes.
Now... Take away this freedom to converge our eyes on an independently selected part of the scene, both sharply and without doubling, and what do you have? An unrealistic viewing experience that is literally dictated by the filmmaker. Your eyes can no longer wander about freely within the subject space. They instead have to search for and locate the only portion of the scene where the eyes can make sense of anything - at the intersection of lens convergence.
If we use selective focus, whether as it has traditionally been used with 2D films or expressly as an aid to hiding the unsavory artifacts of convergence, every person in the audience is forced to attend to that portion of the frame where the butterfly resides, for example, and nowhere else. Gone is the ability to even momentarily converge our eyes on objects that lie elsewhere along the Z-axis. In my opinion, this utterly destroys the sense of realism that 3D filmmaking purports, because it very loudly screams "this is synthetic 3D" and "you must converge your eyes where I tell you to converge them." What's needed is the humility to relinquish control of focus and convergence altogether.
I've got what I believe are some truly viable ideas for how to achieve extreme DoF and zero convergence without suffering several constraints that come to mind, but first...
Can someone explain why it's beneficial to shoot with convergence in the first place?
Thanks!
Mike
I use the hand held ones:
