Production Presentation: (06/03/25)
Production Presentation Link: https://ulster-my.sharepoint.com.mcas.ms/:p:/r/personal/weng-c1_ulster_ac_uk/_layouts/15/Doc.aspx?sourcedoc=%7B482D9ED3-3B14-4565-8AE0-B5562D126971%7D&file=Who-Donut%20-%20Production%20Presentation.pptx&action=edit&mobileredirect=true
- Feedback: We lacked substantially in terms of animation and finalised character rigs at this point in the project. However, we were assured that we would be able to catch up in time.
- Analysis: What I took from this advice was that although we were behind on our deadlines, if we committed ourselves fully to this project – then we could get it done on time.
- Reflection: Overall, I wasn’t happy with where we were at but I felt like could plan more group meetings and deadlines in future. Looking ever forward not back.
3D Character Model Creation Pipeline
It took about 58 DRAFTS of my character model to finally get my design and rigging all in order. This was a testament to my ambition to my character look and move smoothly for our final submission. For simplicity and brevity, I discuss the evolution of my model below in terms of various “versions” throughout it’s development.
DESIGNING
DESIGN – Version 1
This model was used to block out the basic proportions and parts of my model. It informed the level of detail in my next draft.
DESIGN – Version 2
We see the basic elements and design of my character coming together and his general design becoming clearer and less blocky.
DESIGN – Version 3
With detailed hands, and subdivision surface modifiers applied – my character is beginning to look far closer to the final design.
DESIGN- Version 4
PROBLEM: Initially the idea was to have two arms and rig the right arm such that it could curl around into a mug shaped handle. My initial idea was to just add more bones along that arm. However, the automatic weights didn’t allow the handle to look smooth. So then I considered adjusting the weights myself (which I did with great success as you’ll see later when it comes to the armature) but in this specific instance, the bones still looked very “janky” and didn’t smoothly curl into the position I wanted.
SOLUTION: By deleting the right arm, I was able to create a handle that much more smoothly matched my intended effect. I would have liked to keep the hand but despite my group’s best efforts we didn’t seem to be able to join the hand to the handle in the time we had to get the model done. However, I enjoy the design we were able to create – it’s asymmetry makes it look organic rather than simply mirrored topology and rigging. It was a valuable learning experience, teaching me to think ahead in future to the intended purpose of my rig.
DESIGN- Version 5
PROBLEM: The final problem to be dealt with was that the model’s sunglasses did not cleanly and smoothly reflect light as I wanted them to. This was because of the topology of their mesh seen below. The tris at the center (and the way the quads do not uniformly emanate from them) is what caused this effect.
SOLUTION: It was fixed by deleting these faces and instead using the “grid fill” effect which also had the bonus of eliminating those unnecessary tris. The wireframe is rendered red to show the geometry clearly.
The desired effect is shown below.
DESIGN- Version 6
Below is the final design of my model with controllers and armature visible. Overall, while there were various problems I encountered throughout the process of designing my model – I am ultimately happy with how it panned out.
RIGGING:
I also discuss the evolution of my model’s rig below in terms of various “versions” throughout it’s development.
RIG – Version 1
The first rig I created was relatively fine.
PROBLEM: However, when I parented my meshes to it with automatic weights – the weights were all wrong, with different bones in the armature deforming vertices I didn’t want them to. Below is a video quite comically illustrating this.
SOLUTION: Problems with automatic weighting are quite common in 3D rigging, especially when the armature is deforming multiple different meshes like in mine. To correct the problem I sought the FEEDBACK of my tutors Alec & Rachel. They advised me on how to use weight painting to target specific vertex groups to specific bones.
BEFORE (Poor weighting – forearm bone is weighted too much to the main body)
AFTER (Forearm bone now adequately weighted to it’s portion of the mesh)
Overall the results were fantastic! Below is a video of me moving the various bones around to illustrate.
RIG – Version 2
This version involved adding controllers with shape keys connected to them via. drivers.
PROBLEM: Although I found the key shape system rewarding – it was also at times very confusing.
SOLUTION: FEEDBACK form my tutors and teammates however, informed me to keep my naming conventions specific and easy to read. And THAT is what helped me the most. Below are examples of the consistency of my naming conventions. (Left-To-Right: Outliner, Vertex Groups, Shape Keys)
(“CC” stands for “Coffee Cup”)
Below is a short video, taken after my controllers were all completed, that I sent my teammates to explain to them how my rig deformed and how to use it’s controllers for different facial expressions.
RIG – Version 3
In this final version I ironed out minor problems:
PROBLEM: I was unsure of how to affix the cane to my model’s armature. So researched this video on “child of” constraints.
SOLUTION: I used a “child of” constraint, targeted it to a finger bone, and set the influence to 100%.
PROBLEM: The arm of my model twisted on itself strangely.
SOLUTION: I asked my tutor for FEEDBACK and they advised I add another bone for the wrist to the armature to help it deform.
BEFORE (sharp twist)
AFTER (smoother twist)
Animation Process
Prep Work:
Below is my whiteboard sketches used to work out the timing of my animation using a standard 25 frames per second for my 13 seconds of animation which came out to 325 frames in total. The frame numbers circled in red are where I knew my camera shots would need to be along my timeline with reference to the 3D previs. First I placed the cameras, then the characters, THEN began animating.
Layout & Blocking:
SUMMARY: The basic layout and summary of my animation portion is the coffee cup partner and baguette detective standing in front of the jam donut’s body after tasting its blood-jam and arguing over whether the jam is strawberry or raspberry as the jam donut slowly regains consciousness.
Animating – Keyframes:
SHOT 1 – Baguette arguing blood is strawberry
SHOT 2 – Coffee Cup arguing blood is raspberry
SHOT 3 – Baguette getting more annoyed
SHOT 4 – The two face off
SHOT 5 – Jam Donut wakes up
SHOT 6 – wide shot to show all 3 characters
SHOT 7 – zoom shot to show characters’ realising he woke up
Animating – FEEDBACK FROM MIKE (PROBLEMS & SOLUTION)
Later on during the animation process I asked Mike to watch an early version of my animation and give advice that I could implement. Here are problems addressed by his FEEDBACK and solutions I came to with reference to the relevant 12 PRINCIPLES OF ANIMATION (with video cropped showing the solutions in actions).
For RESEARCH I watched this video on the 12 principles:
And I also referenced the following media to study movement and form:
- “Touch of Evil” by director Orson Welles – I referenced this film for it’s influence over the film noir genre we were so interested in.
- “The Three Stooges” – For its emphasis on physical comedy and existence in the same general era as film noir
SHOT 1:
PROBLEM: The baguette doesn’t look like he’s leaning in – he looks like he’s gliding downward.
SOLUTION: I made the baguette lean back before looming forward and made more use of his body’s different bones to create a sense of ANTICIPATION which makes his action seem more organic & dynamic.
SHOT 2:
PROBLEM: The cane is rising too slowly. The moustache is just moving side-to-side so it doesn’t look like he’s talking.
SOLUTION: I made the rise and fall of the cane happen much more quickly. I also made the moustache move up and down as well as side-to-side to indicate that he’s talking. The various facial expression going on along with the slam of the cane all constitute an example of the principle of OVERLAPPING ACTION used to make the shot more interesting for the audience.
SHOT 3:
PROBLEM: Mouth driver not present – needs added in.
SOLUTION: Simple solution here was to make the mouth move and generally match the pattern of human speech. The extreme proportions used here to show the immense anger of the baguette also constitute a comical use of the principle of EXAGGERATION.
SHOT 4:
PROBLEM: Figures jump to centre with no sense of anticipation.
SOLUTION: I added a kind of performative ANTICIPATION to their jump with both of them leaning back and their hats almost falling off backwards. Also, I added in some of the principle of FOLLOW-THROUGH by having the hat move forward as they lurch forward and having the baguettes hand wobble a bit even after they’ve stopped. This attention to TIMING is very important for the scene to work.
SHOT 5:
PROBLEM: The donut getting up is too looks too smooth and floaty and just doesn’t look right.
SOLUTION: Have the donut sit up, then roll on it’s back to allow it to do a superhero stand-up. There’s double ANTICIPATION here, with the donut rolling forward then backward before jumping up. Another notable principle here is the ARC that the donut jumps up in. There’s also some SLOW IN & SLOW OUT principle here with the donut speeding up in the middle of the jump only for that speed to fade out as he lands.
SHOT 6:
PROBLEM: Too stationary, not enough going on.
SOLUTION: More EXAGGERATION & OVERLAPPING ACTION here to add comedy to the scene.
SHOT 7:
PROBLEM: The characters’ need to do something reactive when the camera zooms in to the donut hole.
SOLUTION: The coffee cup and baguette are made to turn to the camera now in shock and surprise. I think the donut’s body framing the faces of the coffee cup and baguette contributes a bit of charm to the scene and adds the principle of APPEAL for the audience.
MY FINAL ANIMATED SECTION
REFLECTION: Overall, I feel very confident amount my final animation despite the struggles and problems I encountered. Although, my timingcoudl have been better and my animation more experimental.
APPROX WORD COUNT: 1706