Exploring the 3rd Dimension – Part 1

Wednesday, May 11th, 2011

I’ve recently began my study of 3D. At first, I must admit, I was a bit resistant to 3D. I probably used all of the cliche arguments. I am, now, warming up to it for a few reasons. First, I’m discovering that there is truly an art to 3D cinema, and that it’s not just a gimmick. There are quite a few good and bad examples of 3D out there. Having seen enough of both now, I can truly say that I do enjoy watching good 3D!

The second reason I’ve warmed up to 3D is because I believe it’s here to stay. From a financial standpoint, any movie released in 3D makes more money. And the people that make decisions on what we shoot are the money people. So if you can’t fight ‘em, join ‘em! With that, I’ve decided to embrace the new trend and learn as much as I can about it.

The first thing I did was participate in the ICG Stereoscopic Workshop at the Sony 3D Technology Center. This workshop is free to members of Local 600 and is well worth the time. It doesn’t focus too much on the technical aspects of 3D, but more on the aesthetics. You learn the vocabulary of 3D… and boy does it open your eyes up to a new world.

Since the workshop I’ve been trying to watch as much 3D as possible. I realize that I need to be able to understand 3D as well as I understand 2D Cinema. Most recently I saw Thor in 3D (which was converted from 2D), Hoodwinked (which we only made it halfway through) and Rio (which I had previously seen with my son in 2D and was excited to compare the experience in 3D). Without making this into a movie review, I’d like to comment on things I’m learning about 3D.

1) Thor & 2D Conversions – these really don’t work for me. It just reads like a bunch of cardboard cutouts placed in space. Of course, all of the CGI is rendered in true 3D… but anything acquired in 2D looks odd. Plus, the lens choices were made for 2D, but when converted, I believe it forces post production stereography to make compromises with inter-axial settings that just don’t work for me. Basically, many over the shoulder shots and even shots where two characters were having a conversation in the front seat of a car, felt like the actors were much further apart than geography would dictate. I hope this trend ends and we start seeing studios make the call to go 3D earlier on in the development of projects. I can’t believe its a good financial decision since it costs $8,000,000 on average to convert a movie and you end up with a sub-par product.

A big challenge that I’m beginning to notice is that extreme wide shots that include people, or animals, just don’t work well in 3D. The people and/or animals begin to look like little models in a diorama. Its really weird, and I saw examples of this in every movie I reference here. I suspect that the filmmakers are choosing an interaxial that is way too big in an effort to make this ultra-wide shot look 3D. Past 30 feet or so, we can’t see 3D. Everything becomes 2D… but we are able to use our knowledge base to determine distance to objects beyond that point. However, I think filmmakers are insisting every shot have as much 3D as possible. So they widen the IO and this changes the relationship of space to size of objects beyond the point we are comfortable with and things begin to look “wrong.”

2) Lighting in Theaters – make sure that the theater turns the overheads off completely. I was in a theater and they left the overheads on, but dim. It wasn’t until about halfway through the movie that I realized why I was having so much trouble watching the film. The overhead was flaring either the left lens of my 3D glasses or my left eye. This was essentially causing an iris mismatch between eyes. This is one of the violations you want to avoid in making, and watching 3D… it makes it very uncomfortable to watch. Once I repositioned myself it was fixed. So if you are in a theater and they don’t turn the overheads off completely… complain! Damn you Regal Cinemas! Where’s an ArcLight when you need one!

3) Hoodwinked (into going to this movie) – besides the movie being pretty bad, so was the 3D. Actually, I take this back… I’m not sure the 3D was that bad… it just didn’t work for this picture. Hoodwinked uses a rather simple style of animation. Definitely not photo-realistic in any way. Feels a bit like the “Far Side” comics… which I have no need to see in 3D either. Probably a bit lax on the textures. So it seemed a bit silly to me to see it in 3D. I suspect it works much better in 2D.

One very interesting thing they tried, which didn’t work for me, was a split screen. This 2D convention just doesn’t work in 3D. Perhaps had they framed the split screen to make it look a bit more “fish bowl” like, but this particular execution doesn’t offer too much encouragement. With that said, I’m not a huge fan of the effect to begin with.

4) Rio – This picture was surprisingly good in both 2D and 3D. I wasn’t surprised it was good, just that it worked well for both! I have to admit, out of the three movies I saw this weekend, I think I learned the most from Rio… but to point out just a couple things…

Dissolves don’t work for me in 3D. Or perhaps the ones I’ve seem are odd. In 2D we are dissolving two, flat images over each other. It works. In 3D, we are dissolving between two spaces with different depth. We could be going from a close up of a picture frame on a night stand, where there’s virtually no depth to the shot, to a wide shot of an expansive room like a church. Ultimately, the dissolve forces these two volumes to fit into each other… its a bit odd! Maybe it just requires some serious storyboarding and planning to be sure the two shots will compliment each other.

And finally, volumetrics. This refers to things in the air that define space like smoke, fog, dust, etc. These are hugely important for defining the 3D world and providing depth cues. However, what I noticed while watching Rio was how amazing volumetric objects looked… not sure if I’m using the correct term here, but I’m referring to things like glass. Objects that have volume, but are transparent or translucent. It was amazing having watch Rio in 2D where I didn’t notice the rendering of glass at all, to 3D where all of a sudden glass just popped! It suddenly had volume and dimension. It was almost as if they couldn’t “dumb” down the glass to match the animation style. Perhaps this is a characteristic of any transparent solid in animation… as soon as its given a 3rd dimension it becomes photo-realistic! Regardless, incorporating glass and reflective surfaces into 3D in both animation and live action is a must!

More to come in Part2…

Leave a Reply

You must be logged in to post a comment.