Production Diary: Testing & Prototyping
Updated: Dec 6, 2022
Briefing
This week we were introduced to the module and the three tasks (which we are currently referring to as "Challenges") we will be completing over the course of the upcoming 11 weeks. The challenges are the following:
Challenge One: Lip Syncing and Facial Animation (Weeks 2-4) - We will be creating a lip sync and facial animation through class demonstrations. The submission for C1 will be an 'enhanced playblast' of the facial lip sync animation with audio.
Challenge Two: Prop Modelling (Weeks 2-11) - We will be creating a prop or background element using polygon modelling techniques showcased in the class, with a focus on 'clean' modelling and geometry flow. The submission for C2 will be a show reel 'turnaround' shot of the model I would have created.
Challenge Three: Performance Animation (Weeks 2-11) - For the final Challenge, I will be randomly selecting an audio and I am to complete a full body and facial performance based on that audio. The audio randomly selected was labelled as "Clip (10)" so that is what I shall be referring to the audio as; you can find a copy of the audio below.
The submission document for C3 is an 'enhanced playblast' with audio. Additional rules have been placed to keep artistic authenticity and originality: no looking for the audio, where it originated, any animations that may have used the audio, and only limited to using the rig provided for this task. (Due to not knowing the origin of the clip being used, I will be referring to the character by the name of the character rig being used for C3: "Eleven.")
Upon initially hearing the clip, I had some initial ideas as a reaction:
The character (Eleven) is cooking a meal to the side while leaving a voicemail for his father. I felt as though this would highlight the casual energy of the line to contrast with the absurdity.
Eleven speaks to two gravestones: one for his mother and one for his father. This is to go further with the idea of the father holding the grudge; even in death, he has still not forgiven him. The scene would play as a family argument.
A TV-sitcom with a laugh track playing. Laugh tracks are often used to cue the audience to laugh and break the tension of a scene.
Eleven would be at his mother's funeral wearing the Crocs, while speaking to his father. This would not make enough technical sense due to the audio referring to "still holding a grudge" and the past tense implying the funeral happened quite a long time ago.
I was liking the first two ideas since they highlight the absurdity of the story, and even aid the story to a certain extent. I consulted my teacher for advice. They said I should focus on certain elements of the dialogue more (specifically the "grudge" part) and exaggerate the story of the animation even further:
Eleven is cooking talking to a pot he is cooking, and pulls out a piece of his father's body or object related to him. The implication is "Eleven is cooking his father and is still upset at him for holding the grudge." This initially was both appealing to me in terms of a dark twist, but also I initially felt it was quite gross. I find gore and cannibalism quite unappealing to my tastes in terms of storytelling; so although I find this as a very interesting and appealing idea, I would need to gauge my comfort levels for focusing on this type of story for the 10+ weeks.
Eleven is speaking to an open casket and he does something inappropriately silly. I like this idea quite a bit. I feel as though I can stomach it, and it raises the tension as the audience realises the gravity of the situation, and the tension is broken by the inappropriately silly action occurring. Some ideas for the silly action happening included Eleven's bow tie spinning, wearing a spinning hat often seen on children, etc.
Overall, I am very interested in this module. I think exploring character expressions through body language and facial movement is quite fascinating, and I am excited to learn more.
Additionally we did a KitBash Challenge, where we were given several assets and we were to build a model out of the pieces. I found this one quite challenging because I am used to building characters out of basic shapes in a very abstract/stylised 2D manner rather than using robot parts.
Upon completion I realised that this task was actually a very good way to challenge my creative thinking and developing art unconventionally. I wasn't too proud of my KitBash robot, but I am extremely happy at how much I had fun and was ready to tackle the course.
Challenge 1: 'Predator' Audio Lip sync & Facial Animation
We were given an audio to practice facial animation. The audio was haunting and creepy based on the sound effects and the voice acting. We started the lip sync and it took far less time than I expected; I assumed getting the lip sync animation done would be some of the most difficult animation work of all. So I was really surprised when it only took a few hours.
I actually really enjoyed the mouth movements a lot, something about how the lips moved were really intriguing. However the cheek inflation really was fairly difficult for me to grasp: with my lack of expertise around understanding air pressure and how to pronounce syllables, I was really struggling to apply it well.
Lower Hemisphere:
I moved on to the upper hemisphere, using a reference. I have very much often struggled with translating a reference with human proportions onto inhumanly proportioned characters, so I was not looking forward to this part.
I took back to my research and applied my knowledge from Alessandro Camporota's video on facial expressions and subtle details: from slight flick of eyebrows to the quick stroke of the eyes. I don't particularly think this work was good, but for my first time doing facial animation it was effective.
Facial Animation & Final Work:
Challenge 2: Modelling Tasks
Dice & UV Layout:
To practice our modelling skills, we did some practice. First of which were some dice. I was pretty confident in my work since I did have some experience dice modelling previously in a workshop.
However, we were to make a UVs to make a texture map in photoshop; which I did not have experience in. I did not enjoy this activity due to the complexity of the UVs, and simultaneously I was enjoying learning about new ways to apply textures for the animation industry standard. I think this was due to the fact I was learning about this for the first time. After this, I still did not particularly feel fond of this, but it was important information I appreciated knowing.
LEGO Car:
This modelling task felt so odd. Understanding how building blocks (LEGOs) require a person to place the object on top of each other to create a shape made this task quite illusion breaking when the blocks were being translated into one another. It also felt so unique modelling the building blocks, which I have never experienced before.
Toaster:
The next object we modelled was a toaster. I found this one perhaps the most tedious? I think it was more due to the fact that I struggled a lot with the mirror tool. I tried three separate attempts at the mirror tool due to user error, so a lot of my time with the toaster was redoing the asset from the beginning to the mirror tool at the end.
I did like the EP Curve tool since I did use that previously for my chair modelling last year. I also remember not enjoying that tool due to its unnecessary complexity, so perhaps I might need time with some tools to understand them.
Candle:
Candle was an organic object, one of the first I was actually modelling. The vertices weren't to be in line with each other, but rather more spread apart since it makes it more organic. In this, I realised the importance of quadrupeds and why five-sided and three-sided polygons are harmful to the model as they often cause certain points of the model to contort in ways that break the illusion of an asset.
Challenge 2: Set Build & Prop Model - Casket
For our prop build, we had to make an asset of an object our character would be interacting with in the scene. I wasn't a fan of the first final idea developed with cooking, so I decided on the character interacting with a casket.
Upon developing this idea, I thought about shifting to a shovel due to a change in my ideas, but returned back to the casket after realising the complications. For further discussion on this, I process this further in my idea development for Challenge 3.
I decided against a coffin due to its very grim and unappealing shape. I felt as though my character would want his father's body to rest comfortably and peacefully and treat him fairly; even though there may still be some unresolved tensions between them. I found a reference and started working.
For this, I did the bottom half simultaneous with the top half. I had a fairly reasonable time with the bottom half where the father would lay, whereas I took the most time on the top half. I really struggled with the little dip at the middle of the casket.
I wanted to make the casket door two separate halves like it was in my reference. However I made the error of bevelling before duplicating my object, so the edges from the flat side were deleted, and when smoothing, the side would collapse into each other. The Insert Edge Loop tool didn't work since there wasn't any completed edges on that face for the Edge Loop tool to follow. So I manually added each one in with the multicut tool, which became a tedious process. I wasn't sure how to accomplish it, but after much trial and error, it was done.
Another aspect I felt I wasn't confident in was the linens and pillows. Since the materials were a lot looser, I would have to convey that same sense of light and fluffiness. I spent much time trial and error the different linens and cloths but for the sake of time I settled on my final. It is effective enough, but not extremely great.
I was still not confident in my UV skills, so I used lambert shader to add colour to my Casket.
For my set, I used several assets from CGTrader: A flower pot with tulips, and a luxury building. I chose this building specifically because I wanted to highlight the absurdity of the situation, as a rich country man is speaking to his father. I also wanted to showcase nature in the story since the audio had birds chirping and bugs making sounds, so I decided to have it outside the building on the balcony.
The flower pot with the flowers didn't come with a colour material so I added it in myself. I specifically chose white flowers because of their symbolism of purity; and in my cultural background, white is a colour of mourning.
Challenge 3: Performance Animation - 'Crocs' Audio
Idea Development:
For my idea, I decided on the second developed idea formed: where the son is talking to the casket. I decided against "doing something silly" because that felt open endedly vague. I decided I wanted more of the tension through actions while the dialogue contrasts with its ridiculousness.
Midway through idea development I came up with this idea of using a different perspective. Instead of a side view, I would complicate the camera a bit with a low angle view staring up at Eleven in a dug up grave as he's talking to his dad's body. I was going to shift my prop to a shovel; however after a review session with my tutor, I realised I was over complicating my idea for no good reason. Although it may look technically interesting, I may become fairly difficult for the audience to understand. "It's not about the story you tell, but how you tell it." So I returned back to my Casket.
Acting for Animation (Reference Footage):
I started off with a bit of exploring the shovel idea, however after recording some reference footage for that, I realised I wasn't really interested in this needlessly complex animation. I felt as though a simpler animation would show my point better.
So I reverted back and made new reference footage. I wanted my character to showcase their grim interaction with their father, and I felt like the story needed to have the character actually interact with an object or do something with their hands while making clear this was a funeral. I made new reference footage using a pen, to represent a flower my character would hold. I was trying to accomplish the final goodbye of a son to his father.
Although I was happy with the body movement, I didn't really like the facial reference. So I redid the reference on its own.
I really liked this reference footage, it added the tired confrontation of my character I felt I wanted to convey.
Facial Animation:
The emotions from the audio are defensive and exasperated. I felt like these words were the final goodbye before the closing of the casket: his final chance to say these words.
Overall, I didn't have as much issue with this as I previously did with Challenge 1's "Predator" audio. The only part I struggled with was the last sentence "Mom would've wanted it that way, she was casual!" which was spoken very quickly causing me to struggle to find very specific key frames and breakdowns to apply. I also found "Casual" difficult to portray, but overall I am very pleased with how it turned out.
Below is a shot progression of my lip sync and facial animation.
Blocking:
Blocking was the thing that took the most time out of all my work. I often have to spend the most time of any animation project blocking. I spend a lot less time on splining because if the foundation of the animation is strong, then the splining will be a lot less horrendous.
It was at this point I realised several issues with my work and footage. The reference footage for my facial animation wasn't in sync with certain parts of my body reference footage. In my facial footage, my character looks to the side as I did in my reference which is due to a head shake. Whereas in my body reference footage, I was shaking my head to a smaller extent.
To make blocking easier, I attempted an "Every Sixth Keyframe" approach. I made the key poses and important breakdowns, then made further breakdowns and inbetweens for certain movements every sixth keyframe. This gave me a really great foundation to work with, and developed my work a lot better than I expected
Below is a shot Progression of my Blocking for Animation Performance.
Splining:
Splining didn't take much time. I knew most of my time would be spent polishing, so I did some quick splinings applying what I knew about anticipation, settle and additional animation principles.
I attempted a variety of camera angles (which you can find in my shot progression below) which I liked but I found the most effective angle was the same angle I used for my reference.
Upon adding the rose and doing splining, I came across a happy accident. Initially when my character closes his eyes, it was meant to be a moment of thinking. Quietly deciding his next words to say to his dad. When I added the flower, the hand was far too close to the head. I was about to correct it when I saw it playblasted. The positioning made it look like the character was sniffing the flower deeply, which was something I absolutely fell in love with. So I didn't change anything and kept that in my final animation.
Polishing:
Polishing was fairly difficult. I knew what needed to be improved, but I didn't really know how to improve it. I attempted to do more animation principles in the Graph Editor, but there wasn't really much to say.
Below is my final work, and I am very pleased with how it turned out.
Comments