Month: August 2022

Continuing with the character modelling process, this week, I decided to focus on the third character, who made no directly mentioned location, giving me many architectural areas to have creative freedom with.

Scouring interest for more referential ideas, I came across this model of a building which really inspired me with its overall shape and details. Particularly the guttering and off-centred balcony placement. I particularly wanted to accentuate height and detail within these and apply them to my own stylistic voice and visual identity.

Adapting these ideas into drawing, I wanted to give this character a kit of pipes and gutterings that could feel like structural support systems that will add contrasting details within the toon shader. I also wanted to create a light brickwork pattern within this to break up the plain white painted walls of the building. Due to the length of the design of this building, I felt I had a lot of room to work with facial expressions and intend to utilise the upper space to move the eyes and the eyebrows. As stated by Tinwell, a majority of facial expression is conveyed in the upper regions of the face, so this gives me plenty of opportunity to explore that (2015).

Beginning the modelling process, I used similar methods to those used with the first model to try and make a distinct link visually between the two buildings. Due to the fact that the interviewee this character is representing states that it was his birthday on the day of lockdown, I wanted to model him wearing a birthday hat to accentuate and play around with the visual ideas involved with the spoken content.

Using the same prior methods with the tile modelling and the toon shader application, I feel I was able to get a neat and effective tile representation with this method and have carved an effective way of creating tiles roofs which will speed up the modelling process later as this stage of production progresses. Reflecting on the current state of the model, I feel the legs are too thick and short and do not fully embody the length I was aiming to achieve within my concept art that matches the longer main ‘body’ of the building. This length in the legs will also aid the production of walk cycles later on as it provides more space for effective mesh deformation in the knee bends, easing the process of skin weighting with IK handles.

Since contextually, it is this interviewee’s birthday, I began modelling different ‘party’ hats that viewers would associate with birthdays visually. I thought if I incorporated this within the model throughout the entire three-minute film, it would make a subtle reference to the first spoken line of the character and also utilise additional secondary animation.

Initially, my idea was to create a ‘polka’ dot effect by using additional mesh shapes. However, this proved to be quite messy and did not have the right effect when rendering, as it was not explicitly clear that it was intended to be a birthday hat and looked more like a stylised ‘dunce’ hat.

In order to improve the overall audience associative imagery, I found a reference image of a very standard birthday party hat and tried to model around this, especially in terms of the colour scheme. Since the key associative element of the birthday hat seems to be a facial ‘spring’ or a pompom, I tried to replicate this by using a single pink sphere.

Below showcases the finished model, with an extra door added, which could potentially work like an additional ‘mouth’, so to speak, to create a more interesting overall visual representation of speech, and utilise the anthropomorphic elements of the design in a way that could make more unique expressions.

Reflectively looking at the model as a whole, I feel it works stylistically as it allowed me to explore my own stylistic ideas without a specific location contextual design. This being said, I feel the simplicity of it will blend and match the other models created so far despite this different approach. During the process of modelling, however, I felt I could have potentially pushed the detail more with the brickwork and potentially looked into the use of ‘bump maps’ which are “greyscale maps that add simulated height or depth to a model’s surface” (Chopine, 2011, p. 160). Utilizing this, I feel I could create a more overall detailed model, which would benefit from less extensive lighting to give the impression of detail. However, once again, the film aims to remain simplistic in visual design to reduce render times and also create more focus on contextual dialogue and narrative over aesthetic presentation.

References

Tinwell, A. (2015) The Uncanny Valley in Games and Animation. Boca Raton, Florida: CRC Press.

Chopine, A. (2011) 3D Art Essentials: The Fundamentals of 3D Modelling and Animation. Focal Press.

In order to stay more organised with the modelling process going forward, I created a sheet in which I can keep track of all the different models I need to complete and their progress.

Contextually within the film’s interviewee subjects, the second (Marianna) was a resident of Canterbury during the coronavirus pandemic, and aesthetically I really wanted to encapsulate the architectural integrity of the local and add this to the design. Looking at different buildings within canterbury, many of the historic ones appeared to be very Tudor in external appearance, with dark, heavy wooden beams adding a lot of detail. Another thing I noticed when looking at the below images in detail is how imperfect and crooked the buildings are, with ‘layers’ and overhanging upper floors.

When adapting these images into an animated character, I tried to resemble the ‘crossbeam’ patterns associated with this particular style of building to add depth to the character I feel would translate well into 3D space. I also felt these beams would be a very interesting way to visually explore expression within the character design, such as eyebrow manipulation. The detail within this drawing, I feel, will also work very well with the toon shader as it will be outlined very distinctly against the move colourless, pale and simplistic base of the model.

Utilising the tutorial from last week, I felt it helped create a solid base more distinctly for this style of building, as the shape of Tudor buildings seems inherently ‘wonky’ and inconsistent. I also felt this tutorial’s use of the roof was exactly the sort of shape and style I wanted to accentuate within this model.

Completing this more detailed and layered roof, alongside a stronger ‘crossbeam’ pattern, I was able to make a more efficient method of layering the tiles so there was not as much mesh collision, which will lead to a much cleaner render overall. Especially with the application of the aitoon shader.

An important part of this model, in particular, was to preserve the detail in the tiles within the toon shader by applying a texture to them individually. This also allowed for specific individual control of colour shade and shadow/ specular contrast.

Similarly, with the last model, I wanted to ensure the ‘mouth’ or door would operate in a way that would effectively portray the ‘jaw bounce’ process of lip sync animation. Considering techniques for rigging this, later on, I will need to ensure that there is a clean pivotal translation that I can parent to a NURBS curve or circle that will have rotational control.

Fleshing and filling out the rest of the details, the eyes are a considerably essential aspect of the animation process, therefore felt important to design these carefully. As tudor-style windows traditionally have a central beam for support purposes, so I wanted to play with this to act as an ‘eyelid’. That way, facial animation can get a more heightened level of detail across.

The next important consideration was the legs; for this model, I wanted to create very small legs that would contrast with the overly detailed ones and add a humourous and stylistic element to them.

In an attempt to add further roof detailing that will mirror that of the previous models, I wanted to add a feature that could also accentuate secondary animation details to the movement to enhance the overall effect and believability of motion. In doing this, I created a weather vane as I felt that fit the overall design and time period contextual use.

Below presented the completed detailed model, which I feel proves much more successful than the last as it seems to encapsulate locational-specific aspects that help set the model and the character contextually. I also feel this fits stylistically very well with the overall aesthetic of the film. With the ‘overhanging’ layered aspect of the design, I also feel this fits the 3D animated form in a way that did not work as well as a 2D concept drawing. However, Improving upon this, I feel I could have added more details within the windows, as there is a very distinct wood detailing within Tudor-style windows. Especially windows that overhang outside the walls and also have more details around them.

Next Weeks Goals

• Design Character 3
• Model Character 3

Mood Boards and Art

Beginning the process of this character design, I wanted to achieve a sort of imperfect, ‘crooked’ type of look that did not appear too dishevelled. Turning to interest, I found two references in which I really aesthetically appreciated the way they had a specific charm and character to them, colour-wise but also shape-wise. Specifically looking at the roof tiling, I really wanted to explore how to create this effect in 3D in a more detailed way to counter-balance the simplicity of the overall ‘facial’ design.

As shown below in my character concepts, I explored facial expression translated into anthropomorphized housing, using the door to effectively replicate the mouth by placing it horizontally and trying to place a window to give the impression of a nose. A key stylistic choice I really want to push within the realms of my film, and in particular, this character design, is this imperfect element of architectural structures and how they replicate the imperfection of humans. To accentuate this for this particular character, I want to create a messy tiled look to the roof to implicate ‘messy hair’ and slanted windows to indicate imperfect noses. The crooked chimney, I feel, also adds an element of stylistic impression, and I want to utilise chimney smoke to highlight expression in the characters (e.g. puffing smoke when angry).

3D Modelling

Beginning to translate this model into 3D space, I tried to start trying to figure out proportionally how I would balance the body and the legs in a way that accentuated stylisation and behaved unrealistically but within the realms of realistic previewing. I also feel this aesthetic choice allows space for interesting weight portrayal within the actual character animation.

Below indicated the first application of the toon shader, in which I was trying to see: a) how it would work as a texture going forward for all the individual characters.

B) How, when using the shader, I would accentuate lines within the combined mesh.

I found when separating different elements of the mesh into groups, I was able to retain the detailed information of each object and have them clearer outlined to build more detail within the model without sacrificing the flatter, shaded look.

When beginning the process of modelling the roof, I was struggling to find an effective method to best place the tiles in a way that did not create mesh collisions and, overall, really wanted to push for the detail in this area. Researching tutorials on youtube, I found a very helpful one that illustrated a clear way to create that more detailed but simplistically stylised look I was aiming to achieve exactly.

Applying this method going forward, I was able to create a more effective and detailed roof which looked exactly how I wanted aesthetically and helped me create a method to apply to the other 5 models later on down in the production pipeline.

It felt essential also to add details which would really sell the model as a house, such as chimneys and TV antenna. I also feel with the addition of these objects, I will potentially be able to explore additional secondary animation later on that will help ‘flourish’ the character’s movements (Such as the antenna flicking etc).

An aspect that felt incredibly important to consider early on within this character creation process was understanding exactly how the door would work in place of a mouth. Recalling rigging techniques I have previously learned, I froze the transforms, changed the pivot location and tried to replicate how the mouth would operate when preparing for spoken mimesis.

Below showcases the completed model, which I think reflectively works well aesthetically for what I want to achieve for my film. However, in terms of modelling techniques, I think in future, I will try to create a neater system for tile placement so that it can look slightly more orderly with less collision. I also feel I would add more brick detailing, given the time to try and flesh out a more three-dimensional texture.

Improved Colours:

Retrospectively looking back on the colour scheme of the model, it indicated several issues as it relied solely on the colours being projected from the lights in the scene. I wanted to stick with a blue thematic colour scheme for this model so that it contrasts with the environment and the other characters to be modelled later on.

Goals for Next Week:

• Design Character 2
• Model Character 2

The Animatic and storyboarding process of an animated film, arguably, is the most important step, as a visual piece of narrative work has less value when there is no purpose behind its presentation. Considering this within documentary contexts, since the ’narrative’ displays factual accounts I felt it important to not just replicate real life but really utilise the animated form in a way that can express metaphoric subjectivity. As stated in Storyboarding: A critical History by Pallant and Price, storyboards in the contemporary age assist massively with understanding continuity editing, and it is essential to understand narrative aspects in my film, and how these different pieces of dialogue will stitch together in a meaningful way that can be visually interlinked.

Going forward Into the following weeks, I aim to complete my previsualsation as soon as I can as ”Previs presents the most accurate vision possible with resources available” (Pallant and Price, 2015, Pp. 152). This will also give me a stronger idea of any additional assets I will need following the list I have created below.

Asset Modelling List

Reference

. Pallant, C and Price, S. (2015) Storyboarding: a Critical History. Basingstoke, Hampshire: Palgrave Macmillan. .