Post 9

    For this week, I got a bit side tracked with what I want to call the “side quests” of my animation demo reel. Instead of my third idea of a quick animation with a medium camera shot from the waist up of a female character looking determined as her familiar (cat) hops on her shoulder from above, I decided to do a facial expression test instead. I’m using the “Moisha” rig because of her many face clusters that allow me to manipulate her skin to move in a more organic way. I wanted a theme for my facial expression test to establish a personality for Moisha, so I referenced the flirty poses as well as the eyerolls and hair flips of a classic “mean girl”. For now, I turned off the visibility for the hair and extra accessories to not get overwhelmed by the details. I think I’ll be working until near completion in segments, one segment per expression. This is what I have so far for the first expression.

    The second “side quest” that I began but wanted in my demo reel anyway is creating a metahuman from a face scan and making an animation from it. I wanted to keep it simple but show off that I could do what a possible future employer would ask me to do. I decided to do a basketball scene with various dribbles, passes, and shots. I asked Dylan Robertson-Figaniak if I could scan his face using the free trial of an app called polycam and mocap his likeness. I sat him down in a library study room and set up a video for him to watch so his eyes wouldn’t track me as I went around his face taking 200+ photos. I found that any less than 200 caused inaccurate face deformations and more than 200 was just unnecessary and caused a long processing time. 

    In the motion capture room, I suited him up and gave him hand gloves so that I could also have finger tracking to remove the guesswork of how fingers move when it dribbles a ball. We also had to improvise certain movements because we didn’t use an actual basketball in fear of it bouncing too hard and damaging equipment. Instead we had a cheap plastic ball with no grip, so some actions did include him letting the ball bounce away mid-dribble and continue as if he still had it in his hand. I’ll probably have to counteranimate those actions along with any other issues such as the thumb constantly flipping backwards. So far, I cleaned up all the data so it didn’t have any gaps or swaps.

    I had to troubleshoot how to import the face scan into Unreal Engine, but, once it was in there, it was smooth sailing. Unreal did a pretty good job of tracing out his eyes, mouth, smile lines, and nose. I then imported that into Metahuman Creator to customize his features to look as similar to Dylan as possible. I will have to retarget the mocap, set up cameras, animate the ball, record face tracking of my own face onto his face and so much more. For next week, I'd like to dip back into my "main quest" that I've been avoiding so far while doing little things for these two "side quests".



Comments

Post a Comment

Popular posts from this blog

493 Senior Portfolio Post 1

Image
     For this spring, I am taking the second round of the Senior Portfolio class as well as my last semester in CSU Chico before I graduate. I am excited to take full advantage of what this class has to offer, which is time and feedback. Last semester, I took on way more than I could chew. I made a complicated two person fight, which I would rush before every blog post submission and then have to redo the damage I had done. It was not a total loss though, because there was a lot that I learned about myself regarding my experience, workflow, strengths, and weaknesses. It also made me think about what I wanted to showcase in my demo reel to represent the kind of person that I am.       This semester, I will be focusing on much smaller scenes. I first want to warm up into the semester with a quadruped walk. I’ve been thinking of doing a cat, since I’ve been researching their unique gait of their feet nearly crossing and their back feet perfectly fitting into the track of their front feet.
Read more

Post 2

Image
     This week, I spent time doing the quadruped walk. I first gathered a lot of references of cats walking from the front, side, and top view. I know that all the cat’s feet touch the ground at different times, but I didn’t want to confuse myself. I figured that I would animate the beginning, middle, and end poses first and then offset each step to reflect how cats walk in real life. For the beginning pose on key 1, I had the cat’s front right paw and back left paw stretched forward, while the front left paw and back right paw were stretched back. For the middle pose on key 12, I had the exact opposite feet positioning. Finally, I copied and pasted the beginning pose onto the ending pose on key 25 for a seamless loop.     I did some in-betweens on keys 7 and 19 so that the paws moved in arcs since cats tend to put one foot over the other as they walk. I also added some side-by-side movements in the spine, shoulder, and hips. At the moment, I’m animating solely on 3s, and I’m not going
Read more

Post 4

Image
     For this week, I got a little behind my schedule because I overrode my runway model’s cat walk with some kind of random attack animation that I started but only got two poses down. I was so sad to find that out, because I knew that I foolishly did my walk cycle on one file. I know I shouldn’t have, but I plan to catch up in the next three days, since I don’t work at my job those three days. I’m still fixing up that walk cycle back to its original progress, and I’ll be focusing on implementing the feedback I received from class afterwards.      I also showed up to Mark Pullyblank’s office hours to talk about the walk cycle of a quadruped. I couldn’t quite tell what was wrong with my quadruped walk, because I was looking at very similar references. In that zoom meeting though, it was very eye-opening seeing how Mark used the reference compared to how I did. He drew lines to measure how straight the back was, how far the legs went, and how angled each part of the body was. I also lea
Read more