Masters Project – 1 – The room.
Overview
For my masters project I decided to document the birth of our child Arianne. I wish to expand on the aesthetic look I’m developing as a film maker and incorporate a lot of the methods I employed in both ‘Outward Displays..’ and ‘Language’. I will not be filming the actual childbirth, and most of the works content will be animation. The camera shot footage will mostly represent ourselves in everyday settings, and in an everyday sense. It will be the animations which will transmit most of our emotional states and internal dialogues. Although, due to the nature of montage, (both 2D and 3D editing), the emotional charge of each element will transfer among each other as each element becomes part of a single flow, a singular aesthetic composition. The finished work will be two films to be shown in parallel as an installation.
My plan is for one of the films to not contain any traditional cut or fade editing, and would instead be one single virtual camera shot in a 3D composed space. I intend to create models and cards that the prepared footage (drawings, stills or video) are projected upon, and create transitions upon these physical models. This would avoid the traditional timeline based montage in favour of a space and form based montage, with time becoming the final axis of the works resolution rather than it being its key mechanic upon which the work is created. This idea occurred to me during my research and thoughts as to how the digital media can create a new cinematic language.
I believe this approach would create a film with a more involving emotional narrative for an audience. It will provide a new space and interval (Vertov) based toolset for constructing metaphors and image relationships outside of (and maybe in addition to) the normal full image/scene based montage approach. I also believe that it would feel to an audience that they were seeing an image of thought, a representation of the internal emotional dialogue within the work and within the artist, rather than a collection of actualities strung together within a time frame. By this I mean that the work will exploit the relationship between the virtual and the actual in order to voice an alternative limit-space for narrative/image/movement that follows more closely the sense and structure of emotions in the sense of emotions being assemblages.
The room for the first film.
I am focussing initially on the 3D space that the first film will be shot within. The video footage I am shooting to place within the space I am shooting on GH2 with the flatest image/colour profile I can get from it (Profile is smooth -1 -2 0 -1). The reason for this is related technically to rendering the footage in an environment with a different lighting model, basically the flatter the video footage is the more easily it will fit into the shot aesthetically.
Earlier Lighting model of the space
The room itself is deliberately sparse, and contains very simple geometry. As the film progresses extra geometry will be added and also removed from the room, and I am also thinking of moving and opening walls to reveal the outside (street scenes, garden etc). Initially I intended to have footage playing behind the windows casting shadows into the room, these two early shots have trees from the garden casting flickering shadows into the room. The overhead in terms of rendering time doing this however (as I’m using an indirect lighting model) means it is something I can only do occasionally rather than continuously throughout the film.
Image adjusted from render passes
I intend to render out the film from mental ray using render passes so that I have a lot of control over the final look through compositing. The animated characters, as well as the drawings and video footage playing on geometry in the room, will also have their own passes. As well as giving me control over the aesthetic, this will also allow me greater control over the timing of certain events that happen within the room. I will be able to change the arrival and departure of geometry/characters, their colour, shadows and reflections within a compositing application without having to re-render large sections of the film. Additional render passes I intend to take will be a Z-Depth pass allowing me to alter the cameras focus, and normal & position passes which will allow me to make small changes to the scene lighting post render.
Early rendering of the room at night
The render pass approach, due to it’s flexibility , and the workflow that I’m adopting will allow me to build the films narrative reactively the way that I have done in the past within Nuke. It will allow me to explore techniques of structuring narrative relating to light control on an object/character basis , and in that sense give more an overall cinematic set based improvisation (with exact control over rack focus techniques too). However, this approach is not without its limitations, key among them being render times and the flickering problems of an indirect illumination rendering solution. In order to deal with the basic problem of render flicker, as well as to support my render passes approach more easily I’m following a method similar to the one documented here by Gary Backus.
Final Daylight lighting model
Although geometry will come and go from the room, and the rooms shape itself may change, there will always be the three principle areas of the kitchen, the bedroom and the living space. I did initially plan for the whole shot to be taken from a fixed camera position, I am however thinking of sections of the film where there will be camera movements. There will never be a cut in this first film, but there will be sections that contain camera pans.
During the MSN4 module I started to develop a method for rendering video within a virtually built 3D environment, I have developed this further and can generate much cleaner images and shadows than those early attempts. The geometry that the video is on is still a simple 2D card, but it’s warped more in response to the actual camera angle now. However, this is not a perfect solution, so as a result my control over the camera will be limited by this. I will have to avoid shooting at angles where the illusion (limited as it is) would breakdown. This is not as great a limitation as it seems however as each piece of video geometry will be mainly fixed within small areas of the scene, and it will be quite often linked to geometry that it can move with,as well as warping itself to reflect changes in camera angle.
Just to finish the post here is a link to some footage shot of Baby Arianne when she was 2 weeks old
Meshworks 