Short Film Production Process                      homepage   computer graphics  short film   photo gallery

                                                                            Celambarasan Ramasamy

Plot:

This film tries to capture the sense of fear and mystery that a commando experiences on entering an unfamiliar territory and how ‘innocence’ can be used a deadly weapon in those circumstances. The film is an unbiased portrayal of an incident. The film is planned such that as the shots unfold the audience would experience the same dilemma that a commando would face in that situation. The film leaves a lot of questions unanswered. As a filmmaker I want to setup a foundation and allow the audience to build their own conclusions. The audience is as much a part of the film as the filmmaker is! The film is an exercise in exposing the audience to a logical progression of shots, there by creating an expectation for the things to come and then suddenly breaking that expectation by introducing a twist at the end.

Production Process:

 Before I started with any CG production work, I wanted to make sure that the shots I had in mind would work together as a good narrative. So I put together a rough story reel using some live action clips that I shot with my friends. Figure 9 shows a few screenshots from that reel. I was able to identify certain important elements from the reel that I wanted to replicate in the CG shots. For example, I ended up replicating the lens flares from the flashlights in the CG shot. I was also able to pickup important clues about the various aspects of the first person point of view that I wanted to replicate in the CG production. The reel also inspired me to incorporate various other elements like the staircase and the child’s slippers into the CG version. The captions in the reel, as seen in the last two screen shots, provided me with critical timing information about the various story beats, that I needed while creating the CG shots.  

   Story Reel:

For the film that I had visualized, the music plays a very important role in driving the mood of the film. So, as I was putting together the reel, I purchased some royalty free music tracks and edited them into the reel to get a feel for the final finished version of the film. The shots in the story reel were just placeholders for a different set of shots that I had visualized in my mind. The reel helped to roughly work out the timing of the shots and ensure that the film worked as a cohesive cinematic piece.

As for the CG assets needed for my film, in order to get what I wanted within my limitations of time and resources, I scouted the stock model websites on the Internet searching for “virtual locations” for the film. For example, in order to create the main house where all the action was taking place, I ended up buying two separate models that had features that I liked, and then remodeled and combined them to create the final location in my film. Even though I couldn’t get the exact props and sets that I had in mind, I still had enough choices to get a combination of different things that conveyed the core essence of the shots. Figure 10 shows the original renderings of these models when they were purchased (top two images) and the various stages of the remodeling process after which the final model was created.

Set Construction History:

I also purchased the various other assets like the child and the solider as rigged models from online.

When I started with the production of the film, my first goal was to capture the ambience of the live action shots, in the computer generated shots. I planned to rely on stereoscopic 3D to create a heightened sense of involvement for the audience in the film. The stereoscopic imagery turned out to be a very crucial tool in enhancing the visual story telling in the film.

A few months before I started working on this production, I implemented the stereoscopic shader in Mental ray.

 Click here for the implementation details of the shader and the  MEL Script

The shader had support for an animatable stereo convergence point. This feature along with the support for depth of field enabled me to create shots that mimicked the visual field of the character. In the film, this technique proved really helpful in guiding the audience’s eyes through certain shots. It also helps in a powerful recreation of the first person point of view.

I did some early experiments with trying to recreate the visual field of the character. Some of them worked and some didn’t. Figure 11 shows some of the screenshots from these early tests. The top two images are from one of the very first animation tests in which I got the eye shift transition to work. The bottom image is from a test animation in which I tried to place the stereo cameras into the eye socket of a model, in an attempt to capture the facial features that are sometime a prominent aspect of our visual field. I wanted to see if they would enhance the recreation of the visual field in any way. As it turned out, they didn’t! This was one of those tests that didn’t go as planned.

Early Stereo Tests:

(Requires Red / Cyan glasses)

Motion Capture Clips: