Category: FMP & Thesis

This video compiles the work that I created for my FMP into one video. For further context in how these contrasting stylized videos were used for data collection in regards to my thesis findings, here is the link to the survey [note: data collection and analysis have already been concluded, any responses will not be counted in the study.]

*also note that Video A in the Cowboy Walk Cycle series was shortened to nearly a third of its original length for the purposes of the study, though the full version is included here.

Cut from Page 17

Whilst drawing comparison between Lee and Marek, one may again be reminded of Lee’s insinuation that lowbrow humor and crude artistic style exists mainly due to continually male-targeted content. Marek, a female viewer, seems to be drawn most to programs which revolve around an emotionally-driven narrative, a phenomenon which has been well documented as a preference among female moviegoers (BFI Research and Statistics, 2015), citing Bojack Horseman and Rick and Morty as her favorites. A 2021 study found that in an audience of 20,162,180 Rick and Morty viewers, men accounted for an audience of 73.1%, while women accounted for only 26.9% (AdTargeting, 2021). Comparatively, in the case of Bojack Horseman, 40.3% of the audience is female, while 59.7% are male (StarnGage, n.d.). It may be noted that Bojack Horseman features a multifaceted female deuteragonist, who struggles onscreen with contemporary issues related to gendered politics. Historically, adult animated shows such as Family Guy set women on the back burner, incorporated only to fill the category of “love interest” or “annoying wife”. In addition to the easy conclusion that women are more likely to enjoy content which features other women in a positive light, one may also note that the inclusion of poignant and interpersonal themes in work, without necessarily sacrificing that which has proven successful among male audiences, may open up the gap in gendered audiences of adult animation.

Addendum: Results from the survey are inconclusive as to these remarks, written months before data analysis. As mentioned, this survey focuses on age, rather than gender, and more specific research would be required to draw conclusions in regards to this. That being said, in the video pair in which both characters were female, audience members seemed to have the strongest opinions, citing negative feelings towards one or the other. In the video pair in which one character was female and one was male, audience members preferred the video with a male character.

The purpose of this study was to analyze preference of realism or cartoon stylization with respect to age. Conclusions drawn about how gender of characters and viewers may play into this should, for now, be regarded as coincidential.

My original hypothesis had been that younger adults would be more open-minded to a more cartoonish style and that older audiences would strictly prefer realism in animation. This data shows somewhat different results than what I was expecting, but, excitingly, there is a definite, strong trend depicted here.

The biggest takeaway from this data is that almost all adults prefer realism in animation. Many of the responses I received from the survey indicated that people seemed to think cartoonish animation was annoying, or, as one respondent put it, “goofy”, while realistic animation was more impressive.

Though it’s impossible to say for certain without conducting an even more extensive survey, my theory is that non-animators believe realistic animation requires more talent or is simply more accomplished. There were even some responses that seemed to be attempting to insult me and my own animation skill through phrases like “the female sheriff’s walk was stupid”, thus implying that they believed her over-the-top, ridiculous movement to be unintentional, and that animation in that style indicates a lack of intelligence or skill. It was these kind of responses that helped me the most, as they demonstrated a clear preference towards one or the other without the respondent seeming to understand the implication in their responses (and therefore not skewing the data).

The phone conversation (Dana and Clara) videos were EXTREMELY controversial. I believe this is because very little changes between the two other than the actual animation itself, as both are simply acting out the same dialogue, and it requires the survey takers to think more clearly about the reasoning behind their choice.

What I mean by controversial is this: about half of the responses indicated that Clara was “too realistic” and that made her “creepy”, while about half of the other responses indicated that Dana was “too exaggerated” and that rubbery, silly motion made her creepy. However, in the end, Clara won out yet again, though by less than the realistic videos had other times.

Compare this chart to the others.

Note that in the Cowboy Walk Cycle pair, the contrast between the two styles may have been the sharpest of all, as the realistic version implied dark, serious lighting (which more than half of the responses praised) and a slow, steady walk, while the cartoon one was intentionally very goofy, with floppy, bouncy clothing.

Let’s discuss the findings.

Again, the easiest takeaway here is that adults in general prefer realistic animation. The most surprising turn in the data, however, is that while Gen Z show a strong preference, and Millennials show an even stronger preference, those over the age of 50 seem to have no preference at all.

I’ve got to say, that result did blow my mind a little. It was exactly the opposite of what I expected for that age category.

There are two possible explanations for this outcome.

1. The sample size is too small. Adults over the age of 50 accounted for only 10% of my respondents. Therefore, it could be impossible to determine a real trend, and I’ll have to mention in my paper that this data is inadequate. That being said, I do believe that with SUCH a strong trend in favor of realism for both Gen Z and Millennials, where almost 3/4 respondents preferred realism, it is still somewhat surprising even with a small sample size that there wouldn’t be a trend shown for the older age group.
2. Older adults really have no preference. I’m not sure if I can include this as an “interview” in my thesis, as it’s too personal, but when my father asked me about my findings, he was surprised to hear the results skewed towards realism. He told me that he (a 57 year old) likes to be entertained when he watches animated content and he always found the cartoonish videos the funniest.
   • If this is true for other older adults I have a couple of theories as to why, though they would have to be investigated in an entirely separate study. Perhaps much of my discussion will involve the ways that this can be further researched. My theories are as follows:
     1. Older adults did not grow up with 3D animation, while people under the age of 30 did. To adult viewers my age, Toy Story and the like were a baby movies, and we graduated from that kind of animation to “more mature” media (in the eyes of other younger adults). Because older adults were already in their 30s or so when this technology came around, it never seemed juvenile to them, because they don’t associate it with a younger stage of their lives.
     2. Older adults may have been raising children (us, me) at the time this technology developed. Therefore, they consumed a lot of it with their kids as a family and grew to develop an appreciation for it totally unmarred by association with developmental stages in their own lives.

Anyway. Lots of food for thought. I will be figuring out how to incorporate all these floating theories into my thesis today.

After my discussion with Nigel, I felt much more confident in the current state of my thesis. The ability to get a second pair of eyes on the work and be able to hear from my assessor which areas need a bit of improving is absolutely vital in providing me with the best preparation for submission as well as allowing me to step outside of my own mind and see what makes sense and what doesn’t.

The day before my session with Nigel, I’d closed the opinion survey I was using to collect the data I needed to finish my findings and conclusion, though I hadn’t looked at the results yet. Speaking to Nigel beforehand helped me get a better idea of the direction I want to take my data analysis in.

His biggest overall criticism was that the choice of language I was using was rather emotive and gave an inappropriately opinionated tone. I expected this evaluation to some degree, as I write very quickly and tend to pour too much of my own mind into my words.

My next step is to incorporate the charts I’ve made regarding my data into my methodology, findings, discussion, and conclusion; then revise the full thesis to edit for length and remove subjective phrasing.

I’m more than happy with this shot, I may even use it as the first shot on my showreel. It took me ages to build the set and I’m very proud of it. I have to say that my FMP shots have only gotten better as I’ve gone on- the cowboy shot took me four weeks and looked awful before I revised it, and this one took me about five days and I love it. I feel as if every new shot I create is my new favorite.

For the purposes of my thesis, I think it works well; the ending adds a little lighthearted factor that mirrors the ending of the cartoon version (the tiger shot). I also intentionally gave it a sort of video game style aesthetic, with Street Fighter in mind, which will allow me to further test whether adults favor a video game style in animated, realistic content, or whether it must remind them of such to not be considered juvenile.

When building the set, I also wanted to stick to that Street Fighter environment as much as possible; if it looked too much like real-life Tokyo or Hong Kong that would open the door for critique on consistency to the setting and time period. For this reason I did not use the background plate at all, but kept in the original camera track. However, I did add my own animation to the camera on top of that to fully showcase my animation. As a result, this shot incorporates my motion tracking abilities the least out of any of the shots I’ve created. For the purposes of my FMP, however, this isn’t necessary, as my focus and data collection for this project rests more on the animation styles (realistic vs cartoon) than the motion tracking.

I am now finished with the animation for my FMP. Before I decide it’s done-done, though, I want to fix the shadowmatte a little bit on the Dana shot. Perhaps I’ll revisit some of the others, too. I’ve been tweaking them as I’ve been going and they’re in good shape at least.

Unexpectedly, it took me quite a long time just to settle on the characters I wanted to use for my final shot. I knew that it had to be a fight scene, because the tiger shot, its cartoon counterpart, revolves around martial arts. I decided it should probably be a real fight this time rather than one person practicing or using a punching bag, to maintain interest and add differentiation. I chose the following reference:

Now, the only issue with this reference is that the men are clearly trying carefully not to hurt one another, and therefore are moving with no real force. This is okay, I have faith in my ability to incorporate that, and the slow movement gives me a clearer reference to study from. I decided that the fighter who flips the other one should be painted as the “protagonist”, and set to work casting characters.

Kachujin was an easy choice, she fits the theme and looks realistic while still being outfitted in sort of a fighter-esque costume. But from there, I hit a bit of a wall. Though I had four choices, I was hesitant in picking any of them, deeply considering audience bias and how certain rigs may affect my data in opinion polls.

I didn’t prefer the first samurai, as I assumed that the heavy armor may take away from the animation work itself and prevent me from being able to show off the work that went into it, and unfortunately, the rig for the fourth samurai did not work very well.

The second samurai would be perfect, except for the fact that he’s portraying a villain, and I’m hesitant to cast this character as such due to race-related biases and discomfort that the audience may feel while watching such a clip. Anything that may skew my data due to sociological factors couldn’t be used as trustworthy material.

Which brings us to the real issue I struggled over. I really preferred to cast Eva as the antagonist here, but had to accept, to my dismay, that I simply couldn’t without sacrificing the solidity of my study.

I wanted to use Eva because her costume fits the theme while staying cyberpunk-y, sort of a Street Fighter vibe, thus setting the shot apart from expectations of having to stay culturally and historically accurate. Eva as a rig would be amazing to use, with lots of controls in the hair and every accessory on her body, as opposed to the unmovable accessories and hair on both Kachujin and all the samurai. I wanted so badly to have Eva’s long braid whip around during the fight. Unfortunately, she is simply too sexualized to use for the purposes of this project. For one, I plan to collect data from a broad age range, and her outfit is simply not suitable for children. In the same vein, I fear that this character may sway audience opinion in favor of this shot for reasons other than the animation itself, especially as the fight would involve two women.

I considered for a while using her anyway and using this as an intentional factor in whether adults require animation to appeal to their sensibilities to be considered less juvenile, but it felt like the kind of thing that would require a whole separate, vast study. Perhaps related to what Mariana is studying, which I believe centers around whether audiences of animated films require their female characters to be feminine and conventionally beautiful. It’s an interesting subject worth further consideration.

That left me with Samurai Three.

I came very close to using Samurai Three, but ultimately decided that he, too, must be rejected for the same reasons as Eva: an adult audience may view a woman wrestling a shirtless man from a different lens than they would a faceless, heavily armored guard.

And so it was with much deliberation that I settled back on Samurai One. The good thing about him is that he’s got clear “bad guy” vibes, as well as being objectively “cool”, thus adding interest as well as putting Kachujin in the protagonist seat.

These rigs are both, I assume, Advanced Skeleton, which gives me a lot of flexibility in movement, but unfortunately both lack controls on their hair and accessories, as I’d mentioned earlier. Kachujin herself has no facial rig.

I could rig the accessories myself, but I am nervous about doing so, as the skeletons are already so complicated.

Conclusion: my outlook is hesitant. Also, I expect this shot to take a decent amount of time, as it involves two characters, is required to be realistic, and runs for a full ten seconds. This may be a challenge.

It took me nearly a month to finish the original cartoon cowboy shot, which was the first I’d attempted to do of my FMP, and when it was done, it looked so bad I nearly lost all motivation to continue the work. Up until this past week, during which I’ve been traveling internationally, I’ve deliberately put off revisiting the cowboy shot, as it depressed me so to see myself fail at a walk cycle. But, in between packing bags, it was the perfect project to break up the leftover time, as the pre-built set and tracked shot required no new planning, only extensive animation repair.

As I mentioned previously, I found that starting over was the most efficient option. That choice turned out to be enormously beneficial, and to my relief, I found that I could whip out a walk cycle no sweat in just a few days, as I should expect at my current skill level. I worked closely with the reference to match the timing perfectly and avoid at all costs the “drunk walk” and weird robotic horse of before.

Of course, once the reference is copied correctly, stylization is the easy part. I chose to give Sheriff Amy the bounciest possible hat, a confident march, and a goofy, stretchy bounce in her step. As with Dana- both rigs, may I mention, are by the same creator, Gabriel Salas, and are the best $15 I ever spent- my favorite bit to animate was her hair. The floaty hair gives the final touch to the weight of her walk. I did find myself wishing that her hips/waist stretched as well, to give a nice jelly-like overlapping squash along with the chest stretch.

George, the horse, was a bit more difficult. Though human walk cycles are old (floppy) hat, George is now only the second horse I’ve ever animated, and of course the first was done rather poorly. That being said, this minimal experience gave me the skill I need for smooth sailing on the horsey sea. With the reference at the ready, I focused my efforts on an up-down motion of the shoulders then haunches, in which the hips raise in each passing hoof pose, thus creating an effect of the front of the horse then rear of the horse raising one after the other.

Pleased with the effect, my main goal after discovering this strategy was to smooth out the hoof movements while keeping the weight realistic, and to exaggerate the upward motion of the hips enough to match Sheriff Amy’s bounce while preventing the horse from appearing too unruly. That took some adjustment. The last step was my favorite, of course, the hair.

Once again, the hair drives home the animation of the character itself.

This is not a showstopper piece like the tiger shot or Dana, but it’s achieved what I’d intended, a finished walk cycle for the cartoon cowboy that looks passably decent and fits consistently into the aptitude of my other FMP work while providing a stark contrast between its realistic counterpart. If I could change anything, I’d say that the horse looks worse than Amy. He still looks good, but it’s clear that I’ve only just figured out how to animate a horse, whereas with humans I’m comfortable enough to play around with stylization.

With my flight back home leaving tomorrow, I’ve had little time to work on my FMP since finishing my internship. But with packing, sharing goodbyes, and obtaining the necessary documents done, I decided to fill some extra hours by getting back into that cowboy shot that had frustrated me so much for nearly a month back in July.

I at first attempted to incrementally save vast changes to the animation, but after half a day, decided definitively that the problems lie in the very baseline timing and posing and I simply must start again, this time with a better reference, a better rig, better blocking.

Though the deadline draws ever nearer, I am confident in this decision. I do not believe I could have achieved a satisfactory result with my original characters and animation. In a span of just two hours, I’ve blocked out a new animation with a new reference, and as my tiger shot took only four days, I believe I can eat that lost time by working harder and smarter on the cowboy shot for a better result.

It’s unbelievable to me that I started and finished this in only four days and it’s hands down the best work on not only my FMP but also my showreel…. I feel giddy!!

Again, the wildest thing is how a year ago it took me four weeks to try and fail at a walk cycle, completely new to the software, fumbling around in the darkness.

I can’t come up with a critical analysis of this right now because there’s literally nothing I would change, I’m proud of all of it. I guess if I translate this to analysis, I can say again that I am most proud of my ability to demonstrate the character’s weight and force in each punch as well as deliberately timing the kicks and moment the punching bag hits him. Rigging and creating the bag itself gave me the opportunity to demonstrate weight, arcs, and timing. Between the punching bag and the tiger, there’s a satisfying juxtaposition between heavy and light weighting. It’s really all just good, good stuff. I’ve gotta celebrate this.

Next step is to create the realistic version of this- probably just a loosely related version with the samurai- and then I’ll theoretically be done with my FMP and ready to collect research data, in the final steps of my thesis…. however, what I really should do is go back and make that first cartoon cowboy shot look less abysmal. I’m dreading that because I’m so frustrated with how bad my work is there, but I know for a fact I can do well. I just did this week.

One of my coworkers mentioned to me that this was one of the “better animations he’d seen me do”, but also noted that although it was arguably good animation, it was confusing what the character was supposed to be doing. My reference was of a girl practicing martial arts alone in an empty room, and I’d hoped that his “practicing” would also read as natural. Obviously, though, it didn’t, so I decided to add in the element of a punching bag.

I downloaded the model online, then added the texture and rigged it myself. It’s comprised of four joints, and it was honestly a delight to animate. Though of course it presented the challenge of matching the swing to the punches, it gave me a chance to show off my overlap and weight skills. And speaking of weight, someone also gave me feedback that the punching bag “feels light”, which I completely see and agree with; however; I want to leave it that way because it is satisfying to watch. It can be a lightweight punching bag. I’m a bit too tired to decide what I think it’s filled with, but the best I can explain is it’s like those neon sandy pillows from the 90s. I don’t want to change it now, I’m very happy with the weight and timing of the arcs that I’ve done. I also like that he’s kind of ducking under the swing after the jumping kick, it feels more interactive.

So instead of adding in a comedic feline moment where he sees a cucumber and freaks out, I decided to make it more reliant on the punching bag; otherwise there’d just be too many elements involved. So I think for the ending I’m just gonna have it smack him in the face and he’ll fall down.