For my premise project, My primary goal was to explore different aesthetic animated interviews, with conflicting visuals and tones. The first is a more serious, darker and atmospheric idea, which is a reference to police interrogation rooms. The second is a more abstract and personal stylisation which pushes more performative and fantastical elements of subjective accounts.
Showreel
Below is a short showreel which goes through the different steps to create each project, particurarly breaking down the process of creating the 2.5 D and the layers it required to create.
Critical Reflection
Over the past 10 weeks of this premise project, I was able to get an understanding and apply old methods and learn new ones to see where I want to expand my visual identity as a documentary filmmaker. I think my main successes lay in the personal adaptation of ideas, particularly in the visual stylisation of the 2.5 D tests. However, Since this is only a premise, I can largely see and understand the various aspects that need improvement in order to be a completed film. Firstly, regarding the first interview, I feel that if I had adequate time, I would have improved the animation to be much more lifelike and fitting to the visual surroundings. I particularly feel that the lower body animation was much too stiff at points, and did not flow as naturally as the facial movements. Going forward with this in mind, I will allow more contingency time for these styles of interviews, or opt for a simpler animation style.
The secondary issue I noticed was the quality of the blend between the 2D and 3D assets in the second interview. I feel the AI toon shader creates a much cleaner and smoother line which evokes a much more computer-generated stylised look to it than the rougher lines of my Tiff image. I also feel that the alpha channel cut some of the detail from the edges, causing the lines to look clipped at parts. I think going forward with this idea it will be important to consider aspects such as 3D modelling all aspects of the scene but creating a 2D-looking render, upping the quality of my line drawings, and applying hand-drawn textures to the 3D models using brushes that are consistent with my 2D assets.
Taking these two visual explorations forward, I wish to pursue finding a middle ground between these two ideas and create a coherent and fixed visual identity that uses the more clarified and detailed facial expressions of the first design, with the stylised and abstract nature of the second.
Project Schedule Plan
Below indicated my project schedule plan for creating my short animated documentary film, including time over the summer which will be essential to get a kick start of animatic creation and character modelling attributes. My main goal is to finish all pre-production, modelling and main aspects of rigging by mid-September so I can really focus on character performance from that point on.
I have also laid out and indicated a technical pipeline for myself, so I can consider which order things will need to be completed in for the most efficient time management going forward.
Progressing onto the next few months of work, below highlights the final major project I wish to produce. Within this, I really wish to explore the animated documentary genre and push for an aesthetic uniqueness that helps me shape and adopt a personal style that compliments my drawn works.
https://vimeo.com/725841618/066fdeca6d
Aesthetic Contrasts
During the development and production of my premise project, I explored two visual ideas that derived from different aesthetic inspirations, during the next few weeks I intend to merge my research into the two into one visual identity in which I can adapt the entire piece of work by. As stated earlier in my blog, there are mentions of virtual worlds such as video games within my recorded footage, which I will explore integrating within the film. This may include clips of recorded video game footage that are displayed through a modelled and rendered television screen within my documentary.
Within the next month, I aim to have completed all of my interviews so that I can start to adapt my work into its pre-production stage.
Below is the finished result of the 10 weeks of production, and features two different versions of the short film. The first is my initial plans for the black and white movie aesthetic, which feature post-production editing such as flickering and vignettes to try and emanate the feel of a real 1930s production.
https://vimeo.com/722972206/2ff2e18f27
Colour Version
I have also published a coloured version to accentuate and make clearer the actions and lighting aspects of the rigging and highlight elements that may have been missed within the black and white edit.
https://vimeo.com/722971495/ca414fb7a1
Showreel
Below is a showreel of the different steps and elements of the pipeline which has led to this finished overall result and indicates the building block and steps involved. During this, I summarise the key modelling elements, and preproduction areas such as the animatic, musical development, previsualization and animation.
https://vimeo.com/727405655/7897e88774
Miro Board
In a similar vein to the collaborative module, I decided to use miro to visually track my pipeline and progress as the project went on, so I could get a clear understanding of where I was at each point within production and what my next key step was.
Upon the 10 weeks of production in which I created this project, I would largely say that it was successful in the amount of work produced in the given time period. I think pushing myself to do an unconventional animation style has helped me understand and explore various aspects of keyframing I had not considered when utilising Maya’s interpolative systems that benefit from transitioning between frames. This being said, I think various elements of my project fall flat due to the animation, and it loses its aesthetic consistency at times that can look jarring and slightly rushed. As I have assessed earlier, I wished to add more time contextual dances to my film to create more direct references, however, failed to plan my film effectively around this to incorporate it. I think if I was able to extend my time on the project, I would spend a lot more of my process on the animation section of the pipeline, to create a more consistent stylisation.
I also feel that my limited knowledge of rendering and lighting at times causes issues due to the excessive noise at some parts of my render. While this works stylistically with my piece due to its ‘outdated’ technological context, I feel going forward I will allow myself more contingency time with rendering, as I found I had to settle for certain lower-quality renders due to time constraints. When it came to the rigging process, I was happy with what I achieved, as I was able to create and understand exactly how to make a detachable rig using parent contains. However, In order to push me in the future, I wish to learn more about IK/FK switches as non of my rigs have this included to date. I think this will be an incredibly useful rigging tool later on.
Another key area I want to push in future projects is texturing. As I stated in one of my earlier blog posts in research, I stated how I wanted to explore the incorporation of German Expressionist painted background into my scene. Due to time constraints, I was not able to allocate efficient time to this cause, but I think learning to paint textures onto 2D models will help immensely with projects and the future, and push myself to create a more independent visual identity in my work. To look further into this idea, In the following months, I am going to explore 3D painting within software such as Nomad and Procreate, as well as photoshop, to get a clearer understanding of this style of texturing. This may also extend to learning how to use Z-brush for sculpting purposes.
Considering my main focus of this film was animation, modelling and 3D, I am happy with the soundscape I was able to create with my own abilities. While I had intended to reach out to musicians to help compose the sound design for the film, I, unfortunately, was not able to, and had to primarily make do with freesound.org and my own composition in Garageband. In future, I think I will continue to utilise this tool to help myself audibly communicate ideas with sound students/ composers.
Extended Reading in Relevant Areas
.3ds Max Projects: a Detailed Guide to Modelling, Texturing, Rigging, Animation and Lighting. (2014). 3D total Publishing. Worcester:3dtotal Publishing.
.3D Animation Essentials. (2012) Andy Beane. Indianapolis, Ind: Wiley and Sons.
. Principles of Three-Dimensional Computer Animation: Modeling, Rendering and Animating with 3D Computer Graphics. (1998). Michael O’Rourke. New York: W.W. Norton.
. Character Animation Fundamentals: Developing Skills for 2D and 3D Character Animation. (2011) Steve Roberts. Oxford: CRC Press.
.3D Art Essentials: The Fundamentals of 3D Modeling, Texturing and Animation. (2011) Ami Chopine. Amsterdam; Focal Press.
.Norman McLaren: Between the Frames. (2019) Nichola Dobson. New York: Bloomsbury Academic.
.The Film Work of Norman McLaren. (2006) Terence Dobson. Eastleigh: John Libbey Publishing.
Finalising my 2.5 D animation style tests, I felt to tie the feel of the section together I have created a title card, which reflects and matches the overall scene environment so that I can create a somewhat convincing transition between the title card and animated shot.
Title Addition
The main thing I wanted to achieve with this is exploring a way I could integrate the questions into the scene without using any additional voices. While I am considering leaving my own interviewing voice into the film overall, I think it somewhat takes away from the isolation of the interviewee in a sense, which is the overall tone I am going for. By addressing filmmaking techniques directly within the film, I am taking away from the performativity somewhat, which is something I want to test within my thesis studies. So for this premise project, I really wanted to hone in on the re-enacted aspect of animation. By including this question I am introducing a context for the speech while maintaining an aesthetic identity.
Interview 2 (2.5D Style Premise)
Below is the finalised shot for my secondary interview, which included the rendered animation from the previous week’s posts. I feel the final outcome along with the edit works well enough for the time being to get the idea across. However, I think in future, I would not create a cross fade cut and would create another pan down to the character, in order to make them feel like they are further part of the same scene.
Interview 1
As well as completing the first interview, I was also able to edit together the first interview process, in which I really tried to stress the ‘crime noir’ aspects of the overall tone, and with the addition of the ‘typewriter’ sound and visuals and the flickering lights. I wanted this style of ‘part 1:’ and ‘part 2:’ to potentially compartmentalize the different sections of my documentary and the questions they regard. When all of the footage has been collected, I believe this will be a good way to help shape the narrative focal point that all the interviews stress when put together, and figure out exactly what the message I want to convey is. I think this will be an effective way to express the general topic without trapping each person to specific questions, and let the spotlight fall more on the subjective experience without the audience’s expectation of response.
creating the finalised edit for the submission next week, I will have time to revalue and critically asses my next steps going forward with this project.
The last shot I need to complete for this week is shot 9, which technically confronts me with more issues than before. Due to the fact that this will be the most ‘skilful’ performance I will be animating, I had many thoughts and considerations about how I would display this.
Shot 9
A key part of this shot again was the limbs detaching in the air and re-attaching upon landing. I also had to plan exactly what dance inspiration I wanted for the rest of the shot, as it had to display the most technical skill narratively. Looking further into ballet dance, I thought that a key associative dance move was the pirouette.
Watching the reference video below, I tried to recreate this with my character, however, came across several difficulties. The main issue I had was getting the timing of the spin correct which also coincided with my stylised animated method. The second issue was that It did not allow for dynamic staging, and I struggled to plan the motion around and ‘trapdoor’. This leads me to consider things similar to ballet that demonstrate associations with flexibility and skill, and I began researching gymnastics.
Inital ‘Spin’ Animation
In order to get this look of flexibility, I decided that for staging and comedic purposes, I thought it would be good to make the character do the splits over the trapdoor, and fall in, allowing time for anticipation and comedic initial defiance of gravity (much like in looney tunes). Looking at the image below, I noticed that gymnastics athletes fall into this position with very gracefully and well-controlled steps, so translating this onto my animation I tried to exaggerate the finger-pointing and chin lift to demonstrate that he is proud of his technical ability. However, Trying to keep the comedic element to the piece, I tried to include the earlier ‘Saturday night fever’ reference.
The Splits
Below showcases my finalised animation for this shot in which I really tried to push for frame rate experimentation, as I animated on 1s to accentuate character fluidity, which is why this shot is my most ‘polished’ piece of animation. I used the same technique at the end to disassemble his body as I had previously in shot 8, however, had a more preconceived understanding of how it would work, which I believe made the process a lot fast and smoother. While technically I can pinpoint several issues with the animation itself, such as
Credit Song
Since I really wish to emanate an animated television show with my film, I decided to compose an end-credit song to coincide with the opening music. Since the film is over, I wanted to add a calmer toned-down piece, which is why the number of musical instruments is stripped. I used the same key and musical theme that has been present throughout the entire film, but slowed the tempo and added a glockenspiel, to give it a nursery rhyme association that is remensient of children’s bedtime stories, bidding the viewer fairwell.
Editing
During the stages of editing, I wanted to add extra text that references the fine print at the bottom of film trailers in the present day, but parody this by adding text such as ‘the skeleton men present’ to almost making fun of the format due to the obvious lack of studio presence in my work.
Initial Edit
Below indicated my first initial edit for the film (excluding the title card and end credits) and indicated several areas of improvement going forward. First is that I need to find a way to apply the black and white film filter options as previously explored, to the whole feature while retaining as much of the render quality and contrast as possible. I intend to do this but upping the temperature and tint, as it will bring forward the colours and contrast, making them more apparent with the black and white filter. Secondarily, I need to complete the start and end credits and add them alongside the film itself. I intend to make two copies of the film, one if black and white and one in colour so that the render quality can be retained in one and visual technologically contextual settings can be played with in the other.
One of the key parts of the visual success of this shot will be to try to effectively blend the 2D assets with the 3D ones to make a visually convincing and sustains its aesthetic credibility. In order to create a line-drawn effect within 3D space, I researched Arnold toon shaders, which add this exact effect but retain the 3D element of specular texture. Looking into this I found several tutorials that helped me understand that the Arnold render filter needs to be changed to ‘contour’ and line elements can be changed and adapted within the attribute settings, similarly to aistandard surface materials.
Applying the above tutorial to my 3D character model, The line effect works well for the stylisation of my shot, as it blends it into an aesthetic world in which key elements are outlined. Hence why I did not apply the toon shader to background elements such as clouds, as I really wanted to visually accentuate the foreground and key elements to draw the viewer’s eye.
Ai Toon Shader Render Test
I also added this to modelled elements of the midground and foreground for the same effect. However, where this visual look falls short I believe, is the drastic different line quality between my exported tiff drawings and the 3D models. I think going forward with this project, I will 3D model all of my drawings into 3D space, but apply them to toon shader to get a cleaner, higher quality effect that I was not able to reach its full potential during this premise project’s experimental period. I also think that going forward, I should apply artistic lighting and colouring theory to the rendering, as objects in the front should be more saturated than objects in the background, as this ruins its sense of depth.
Ai Toon Shader Test
Completing the modelling process, due to the specific bit of dialogue which expresses a long car journey, I thought I would directly visually convey the speech within this, creating a more ‘observational’ documentary mode, however, place it in a wildly unconventional world to separate its aesthetic tendencies from the mundane. To upkeep this stylistic direction, I modelled the car assets in a very flat and almost ‘cardboard cut-out’ style to maintain and keep inspiration from the shadow puppetry style Lotte Reiniger researched earlier on in the project. The feel of objects being made from plastic (accentuated object specularity) and ‘ cardboard’ 2d stylisation adds a hyper superficial like aspect to the world that once again stresses a visual landscape that forcefully attempts to push itself into fictional realms while tackling non-fictional subjects.
Modelling a 2.5 D Car
Animated Facial Performance
In order to be time efficient during the animation process, and also exemplify stylistic independence and reference to computer-generated imagery, I decided to try and test using the audio node within Maya MASH editor to convey speech. This would take place of creating and animating a mouth rig on the character for the time being, and allows me to explore the visual identity of my potential film.
Mash node editor
By applying the mash editor node to the mesh, it created a computer-generated audio processing effect which makes my character appear robotic, which I think adds a visual performative element to the mix that forces viewers to acknowledge that the character is fake, in contrast to the voice behind it, which will aid in my thesis survey research into the performative re-creation of subjectivity.
Reference for Eye and Eyebrow Movement
After using the real-life footage of my interviewee, I tried to follow and adapt the expressivity of the eyebrows and eye movements, and apply this to my 3d model adaptation of the concept art below. Due to time constraints, this animation came out very minimal and not overtly accurate due to loss of performance in stylisation. However, this has allowed me space for stylistic considerations going forward with this visual exploration and has led me to decide upon making and adapting more detailed facial rigs for future models that are more reminiscent of the original footage of the interviewees. This will potentially include not using the Mash node editor in future, but creating different ‘square’ mouth shapes that will operate with blend shapes to create the illusion of movement. I am also considering applying 2D draw facial elements to 3D models as a stylistic choice in future, which accentuates ‘paper drawn’ visuals.
Character Concept ArtAdapted Animated Facial Performance
After completing the facial animation, I decided that in order to engage audiences with the visual world-building, I wanted to keep the speech as one continuous shot that does not include shots, which also provided the challenge of aesthetic consistency. Going into next week, I will try to merge and asses these aspects to make the aesthetic design work together.
For this week, I have updated my animation and rendering progress on the shotlist below, indicating my work for this week will be focused on shot 8. Due to the technical aspects of the rig connecting and disconnecting within this shot, I have allowed myself the week to experiment and gauge the best method for achieving the finalised look.
Shot 8
Due to my primary dance influence deriving from ballet, I initially started out researching and watching videos of ballet dancers on youtube to watch how they point and transition between different poses. Due to my more ‘staccato’ animation stylisation, I aimed to take key poses from these videos and translate them in a way that does not require excessive keyframes to make kinetic sense.
Ballet Reference
This was particularly challenging when facing jumps, as the timing and poses had to be very specific to appear convincing on landings while maintaining the poise and elegance of ballet dancers. However, in order to get a better conveyance of weight within the jumps, I animated more frames specifically leading into and out of jumps working as a ‘slow in’ and ‘slow out’ function with the limited animation.
Ballet Reference
The most challenging part of the shot proved to be the last section, where skeleton b’s body falls to pieces into the hole. Due to the way I rigged the character, I could only detach certain limbs such as the arms/ forearms and hands, as the IK placement make it very difficult for me to make the legs detachable. Due to this, I eventually decided that it would be much more effective to place an unrigged model into the scene, and swap them at the frame the skeleton starts to disassemble. in order to get this to look convincing, I had to ensure that the model was in the exact same position as the rig.
The next challenge was making the different bones of the skeleton fall convincingly. In my initial attempts, the timing seemed very off, as I was animating on 2’s and the sense of gravity seemed inconsistent with that of the rest of the scene. Improving upon this, I began to animate the different sections on 1’s, as this created a fluency and sense of weight that did not draw attention to itself.
In order to get a rough idea of how these bones will fall, I took a handful of pens and pencils and filmed me dropping them in slow motion, to get an idea of how they fall. The main thing I noticed was how they rotate sightly but all in the same direction. With the longer limbs such as the arms and the legs, the motion will be very similar and it made the application of this to my animation a lot clearer.
During the fall, however, I wanted to make sure that the head was one of the last things to fall out of frame, to stay focused on the character and his expression.
Film Reel Editing
To add further authentication to the time contextual setting, I wanted to create an introductory film reel roll, to give artistic reference to old film projectors. In order to make this as convincing as possible, I went online and found the sound of a film projector warming up, and laid this with edited effects in da vinic resolve to create a film reel look. Once edited into black and white, with noise grain added, I feel this may look convincing and sets the tone immediately for the film’s aesthetic reference.
Next week’s goals:
Going forward with this technique I will utilise and reference what I have learned from this week’s animation and create a faster and more effective piece in the following week.
. In the next week, I wish to finish shot 9, therefore completing the animation of the film.
. I need to set up and finish at least half of the film renders
Following last week’s shot list creation, I have indicated all of the shots I plan and intend to work on for this week. The shots I will be focusing on are shot 3, 4 and 10, as it will be important to finish and plan both the introductory and final scene of the film so I can get a key understanding of staging, placement and lighting.
Shot List
Style
As used in my previous test animations, I feel that using a style reminiscent of pix-elation will help add stylized and comedic value to my work and suits the overall theme of my film. This also will help me set attainable goals and reduce the amount of time spent in the animation process, making the minute-long film more achievable. A big stylistic inspiration for this is neighbours, as it uses elements of stop motion animation in a different context that adds blunt humour that live-action cannot inherently produce with the same effect. As stated in Norman McLaren: Between the Frames Neighbours led to the creation of “his pioneering stop-motion live-action technique” which is called ‘Pixilation’ to utilise this (Yang, 2020,pp. 167).
I think the ‘staccato’ style to it works really well in creating a particularly visual identity to the piece that does not blend into the world of live-action, drawing attention to itself as a performance due to its inherent abandonment of real-world physics. I wish to create a similar effect that dramatises and differentiates movement in a different way from what is typically expected from CGI animation. I also think a key part of this is due to the control McLaren has over the timing and relationship with the music, and I feel this will be effective in the creation of my own work going forward.
In the video Experimental Film Artist: Norman McLaren (1970), McLaren discusses when working on the film Neighbours, that his use of sound he created animated frames of soundwaves which create different pitches, which are controlled per frame and matched with the film exactly (1970). I think this accentuates that the use of sound with this animated technique is imperative for stylistic endurance.
Another recent piece of animation I noticed takes a similar style is the animated documentary ‘Flee’ which largely seems to have a more limited animation despite its more detailed drawn 2D appearance. There are fewer animated frames that create pauses in movement and facial expressions that minimise movement but do not sacrifice emotional conveyance.
Scene Set-Ups
An important part of the 3D animation pipeline is setting up all of the scenes so that the staging, placement and render settings etc, are consistent between each scene. So my initial starting point was creating a base Maya file for every shot on the shotlist. Going forward, I am aware that doing this process can cause breaks and inconsistencies in a scene where the animation relies on the previous files’ last position, so with that mindset I will ensure that there are different cameras, and files that are saved from the previous shot’s and keyed at the last position.
File Set up
The key element required for every scene is a central, locked camera with the staging measurements in the centre so that everything is in consistent line with each other. The main prop or asset that is required is the ‘trap door’ or black hole in the floor. The key placement of this ‘door’ in shots 3 and shot 10 will be important going forward as they are the start and end frames of movement and points of reference.
Render Set Up
Across all of these scene files, I also tried to make sure the render settings were the same and placed all of the different assets onto different layers that could be combined and manipulated at the end of the post-production compositing stage.
Render Layers
Shots 3 and 5
An idea I have conceptualised is using my N cloth simulation tests to create a curtain opening sequence that adds a further ‘theatrical’ aspect to my project and makes more direct reference to the theatrical setting as well as helping me apply and develop my simulation abilities. By having the curtains drawn open at the camera it introduces the scene in a way that forces its performative status on viewers immediately and slightly breaks the fourth wall by using a curtain to introduce a stage.
Curtain Simulation Testing
Due to the nature of the scene, a Key aspect that I needed to take into account was animating the area lights that acted as spotlights alongside the character. This initially proved difficult as even in IRP render viewports It look a lot of time to understand where it was at each point. In order to attempt to try and match the timing of the animation as much as possible, I keyframed in a stepped stylisation the movement of the lights in the same beat as the skeletal figure to try and ensure it is always illuminated.
Animating the Lights
Below shows a playblast of the finished shot. My main aim with this shot was to understand and adapt how the skeleton’s rig will fall apart and come together under the applications of gravity. For this, I intend to always be aware going forward of the character performance ‘jumps’ as a certain section of the limbs will fall apart and come together. I wish to do this to pay homage to The Skeleton Dance, and also the historic comedic effect of reanimated skeletons that do not have the muscle to hold them together. I also experimented with the idea of varying animated keyframes when I wish to draw the audience’s attention to a certain character. When the character is animated on 1s and 2s, there is normally weight conveyance needed to sell the shots plausibility, and also to draw viewers to the movement. This is why, when skeleton B is dancing in the below shot, Skeleton A is almost stagnant. I think this reflects the work of McLaren on Neighbours, as a similar variation of frame rate is used for different actions.
Shot 4
Shot 4 deals with the only extreme close-ups within my film, therefore requiring the most emotional expression. Due to the fact, that the main characterizing difference between my characters is eyebrow emotional conveyance, I wanted to use this time to stress the difference between the two personalities. Skeleton A is shy and concerned about the upcoming ‘battle’, while Skeleton B is determined and ready to ‘fight’.
Skeleton ASkeleton B
An editing style I wish to use for this section is derived from old western movies, In which the camera uses extreme close-ups to garner characters facial expressions when they are about to duel. An example of this is from the movie The Quick and The Dead (Raimi, 1995). A key point of reference here is the 180-degree rule within the film which underlines the importance of staging and direction, as the position of the skeletons on stage needs to be dictated by the direction they face. I want to accentuate this effect by creating a split-screen, that will exit the screen on the corresponding side of each character. I could potentially use this opportunity to create an “elliptical Cut” which is a “culturally conditioned film convention ” that allows a large jump in action between shots that still works with screen continuity (Brown, 2011, pp.77).
Shot 10
This shot is the last shot of the film and contains the largest camera move in the whole film. Due to this, I had to ensure that the underground environment was fully modelled. An early issue I had was considering the way I was going to model and build an environmental depth without making the scene too mesh-heavy, making my computer and rendering process crash.
What I found worked for this process was getting the different ‘piles’ of bodies, and duplicating them within a specific viewport so they gave the appearance of being far away from each other, expanding the environment and accentuating the scale of failed ‘dance battle’ attempts.
Underground Environment
Below is the play-blasted, unrendered shot which I will begin rendering tonight, which will provide me with key points in lighting, render time and camera movements, as I have considerations of adding post-production motion blur to my piece.
Next week’s focus will be predominantly on Shot 8, as it involves a lot more intensive movement and animation than these previous shots.
Reference
. Experimental Film Artist. (1970). Mclaren, N and Sloan, W. Footscray, Victoria, Australia: Contemporary Arts Media.
.Brown, B. (2011). Cinematography: Theory and Practice: Image-making for Cinematographers and Directors. Oxford: Focal press.
.YANG, D. Norman McLaren: Between the Frames. Canadian Journal of Film Studies, Fall2020, vol. 29 issue 2. Pp. 165-168. DOI: 10.3138/cjfs-2020-0027
