Submission Page – ‘Graham’ – Ma 3D Computer Animation FMP
Final Animation Link:
Rendered Version
Play Blast Version
Show Reel/ Portfolio of work:
Blog Project Category Links:
1. Everything- FMP Catagory: https://mafmpthesis.myblog.arts.ac.uk/category/fmp/
2. Preproduction: https://mafmpthesis.myblog.arts.ac.uk/category/fmp/preproduction/
3. Production: https://mafmpthesis.myblog.arts.ac.uk/category/fmp/production/
Weekly Blog Posts:
Critical Appraisal
My Final Major Project Concept Summary
Firstly, my final major project explored the state of mind of an elderly person in mourning. The narrative slowly develops his two-sided dialogue, even though the audience never see Grace on screen in person. I implement symbology and cinematic techniques to produce information for inferences from the viewer. Directing them to understand that the story is about the loss of their close relationship, and the actions that Graham takes to seek comfort and closure.
My Aims
- Applying the Kuleshov effect into my film: exchanging/ intercutting character expressions with contextual content to provide meaning (for Hitchcock’s facial reactive version of the technique).
- A rendition of Kuleshov’s experiments were applied, altered to apply to my narrative purpose. Based on my premise project research and experimentation, I applied on and off-screen elements, to produce the effect of a presence that is not really there. One created by the character, that produces ambiguous narrative content. Purposefully designed for the viewer to question and created inferences from.
- Within my narrative, my aims consisted of the production of a refined storyline that is easy to follow and most importantly emotionally impactful.
Whether or not I achieved these aims is a matter for others to review as much as myself. In how they understand and experience the cinematics of the film. As much as I may have had reservations about the final outcome, the feedback I received was surprisingly positive.
Many elements could have attributed to the outcome of the animation, though most importantly, the way I was able to cooperate with my computer impacted me the most. Considering the large extend of my file and the lower that ideal playback speed of my computer, I was not fully able to grasp the timing between poses. Also, the significance of pauses between poses was not exercised well enough, especially in the dancing beginning.
Hitchcock’s Kuleshov effect technique was implemented extremely well within my narrative. The concept is simple; therefore, its presence is subconscious in the viewer’s responses to it. The technique was applied to visual content in order to guide attention and the directed message. This, and the dialogue were both very successful in the final outcome. They were easy to produce and construct and my messages proved to be understandable to the audience.
Animation Review
I believe that within the animation part of my project, there could have been many more improvements to refine curves and timing of poses. Poses themselves should have been reviewed further than they were. Considering any part where the character turns and starts walking, those areas are very unnatural. To correct these, I could have used reference footage with key poses on the main frames and then added supporting in betweens, instead of without these. That way I could make my character move with the appropriately looking positions of the body while in motion. The way I animated these parts was by turning the centre of gravity, chest and head. I believe that I missed out on an important part of the turn which was the shoulders.
Furthermore, many other areas, such as the facial expressions and the dancing animations, could have been improved with the purpose of timing. Holding the poses for a number of frames before moving on can have such a strong impact. And allows the audience to view the character’s movements in greater detail and appreciation for purpose.
The issues I faced with the facial animation derived from the rigging and model shape. I deliberate on these aspects further within the appropriate blog posts. However, even after these issues, I have found ways to work around these. For example, the shapes that the Advanced Skeleton rig made for the mouth were not ideal. And I fixed these my redoing the weight painting many times, until I was comfortable enough. Not only this, but I fixed this issue using set driven keys. The effect of this was a customised rig that allowed me to manipulate the mouth to my will while animating. While this did cost me time to edit and correct, I found many ways to help my process. For example, blend shapes that allowed me to standardise the facial expressions into repeatable assets that I could rely on. This was a great help in the long run, to create high quality acting with details in the face, in a short amount of time.
Rigging
Overall, the rigging processes progressed with ease due to the use of Advanced Skeleton. This was my first experience of rigging a face within a project situation. The result of this part of the project was very successful. While, I may have had problems to fix, the process was reasonably easy, and the outcome is usable and acceptable. I have even learned more about blend shapes and set driven keys due to this project. Learning how to master these techniques in order to have the ability to customise my rig to my purpose. If I had to critique the project rigging, a few aspects would need to be pointed out. The foot roll position is not perfect, since the joint should have been a bit further back to where the first toe joint is. This made the toe roll positions look slightly wrong. If you looked closely. When you bend the roll position into an extreme position, the foot lifts too much. As though he is on his tip toes. Furthermore, as I mentioned, the final effect works with the character’s movements, though this aspect could have been more refined.
Lighting
I have gone to great lengths of personal learning, experimentation and practice to create an aesthetically pleasing environment with realistic lighting. It was difficult to decide on the time of day for the setting, and therefore the lighting temperature, however, I chose based on peer feedback and suitability for the scene. Such as the intensity of light that would be necessary in order to see the character. I experimented and learned many new things about how to produce more realistic lighting with Maya settings. Such as exposure, intensity, temperature, angle (softness of edges on directional lights), ray-depth diffuse (bounce lights), indirect lighting (which I found often created fireflies often) light samples, clamping, adaptive sampling, and light temperature. These aspects created a depth of development for my environment realism. In spite of the long rendering times – which would sometimes be 20-50 minutes. The large rendering time caused me to decrease the camera samples while doing testing, that I why my tests are not in the best quality. Also, slowing my workflow progress, nevertheless, it was worth the outcome.
I knew from the beginning of my project that the character would require lighting custom than what can be seen in the environment. Lighting to ensure that he was not in the dark at any important part of the story. Therefore, I provided my character with a lighting rig, 3-point constrained lighting that followed him around. This had to be close enough to not cast light or shadows on the environment around him. This worked immensely well to apply ‘Rembrandt’ lighting to the characters face, creating a more aesthetically pleasing appearance to the character’s cinematography.
Character Model
I am proud of how the character model turned out. However, there were major issues that could have been improved upon, firstly the texturing normal maps seem to not follow the contours of the skin, and the sculpting work I made did not translate either when I baked it either. The eyebrows were made with alpha maps, this turned out very well, however, the hair on his head did not turn out as well. I believe the issue was with the alpha map, since I made this by hand, applying brushstrokes in Photoshop and then converting this into an alpha map. I believe that it did not convert properly, the individual hairs did not stand out from each other enough. Not only this, but the design of the hair models in Maya were not natural enough. Perhaps one mesh with one large stroke of hair would have been simpler. However, I did not do this since I feared my character would not look as realistic, too cartoony in contrast to the rest of my design. Especially the way that I managed to create his eyebrows so well.
In terms of critically reviewing my final major project.
The outcome of the content can be summarised by quantity over quality. While I attempted to maintain a high quality throughout my many environment and character assets. The quality of the animation could have been refined and smoothed out much more if I had spent more time animating rather than adding small details to the narrative for the purpose of effect, story devices and symbology. These elements, such as the framed portraits, the flower décor and the birds at the start. I am passionate about these details, in how they enhance the authenticity of my story. Though, they each took up my time that could have been applied to the animation more. I would generally aim to spend 30 mins each evening on these small projects. I believe the outcome of these elements is the enhancement of cinematic experience for the audience. Based on feedback from viewers, this was successful since they commented on how they were partial to these additions.
Time management and Rendering
The management of my time was essential for this project. I started very early for most stage of my work. Based on previous experiences, this was necessary in order anticipate issues which I inevitably faced. One issue that is note-worthy to mention is the issue with rendering time and the render farm. At the beginning of my project, I had every intention to implement the render farm into my workflow, in order to produce efficiency in my workflow. However, I soon realised that my rendering time was about 55 minutes per frame, once I did everything I could, I managed to cut this down to about 26 minutes. Not only this, when I transferred my file to the remote access computers, the file paths refused to connect despite my many attempts. On top of this, the render farm produced many more failed attempted renders than successful ones. I managed to mostly resolve this issue on my own after much trial and error, however I became delayed in my submission due to this. The outcome required me to cut down on rendering time further by lowering my samples, unfortunately. However, this was necessary to produce rendered frames in a shorter amount of time.
Overall outcome Success or Failure?
This critical review may have more critical analysis, rather than constrictive feedback on my project, however, the fact of the matter is that I have completed a three-minute animation, with a story, character, rig and environment that works within the animation. Not only that, but it is correct and appropriate to the concept. I have achieved the concept that I set out to convey and the story has produced strong emotionally interactive responses. This is great feedback from the viewer’s perspective, that my aims were achieved in spite of my criticisms of my project. Therefore, I personally deem this project with a successful outcome.