Starting off this module I was less stressed than I thought, after completing the first half of the module I felt a lot more comfortable with this assignment. I definitely was a little panicked though, There is a lot about animation that I struggle with and fluidity is definitely one of them. In this blog, I’ll share my first experience of making a lip sync, a jump and a heavy weight lift animation, I will reflect on the struggles, and experiences gained along the way.
On week seven the class jumped right into body mechanics. I found myself still returning to The Animators Survival Guide by Richard Williams to understand the movement during a weight lift animation. (Williams, R. (2001). The Animator’s Survival Kit. 3rd ed. Macmillan). I jumped right into the work and started off by watching a video that Aodhan recommended to use as a reference for a weight lift. I did the usual block out but found that once I switched from constant to linear between keyframes on the graph editor it looked really bad. I knew that it was because of the linear curves but no matter what I did, it just didn’t look natural. It wasn’t until I talked to a guest lecturer that came in, Matthew Hamill that he suggested not to start animating in constant because it messes with time perception. I found personally his advice worked out for me but I know not to limit myself in the future to completely cutting out blockouts. I took a while to decide what kind of weightlift I would animate, I decided on a tire lift and began researching videos to use as references. I could’ve stuck with the class sample but some of the marks are given to creativity and personality shown through the animation so for every animation I wanted to come up with my own idea to show that I could come up with something good on my own and convey everything that I’ve learned. I found another animator who had made a tyre lift animation so I watched that and took notes about their techniques to help me with my process while also watching real references and watching the weight shifts and body mechanics. I ran into a problem with my weight lift animation where I wanted to add a dynamic parent to the tire and the hand control but whenever my hand would rotate to push the tire the tire would rotate as well. I went to Alec with my problem and he suggested making an armature for the tire itself and turning off Dynamic Parent when necessary. Alec made a tutorial video, explaining what I would have to do to do that and I gave it a go. I loved the way it made the tire look really involved in the lift but in my animation, the tire was being pulled towards my snow rig. To fix this I rewatched the reference I had pulled and studied how the tire itself moved, It didn’t slide because of the weight so I went back and changed the animation to look realistic. Once I thought I was finished I got some feedback and with the dynamic parent turned on the hands were constantly going through the tire, even when I tried to change them so I had to revert back to animating the tire. I made a reference video of myself lifting up part of my couch to really understand the position my legs were in, I used the knee controls for the first time after not using them in any of my walk cycles and it really proved useful, theres defiantly a lot of bones and rigs I hadn’t used in the walks that I began learning for the body mechanics assignment. A big chunk of feedback I got about my walks was that I didn’t animate the shoulders at all so that was something I made sure to do this time around. I think that feedback helped tremendously, the shoulder movement is so important when it comes to weight lift so learning from my past mistakes helped me a lot with this assignment. When it came to the tire falling onto the floor I encountered my next dilemma. Because the origin was at the corner of the tyre it wouldn’t rotate without intercepting the floor so to fix this I created three tyres, one with the origin at the bottom left, one with a centre origin and another with an origin to the right. To make it look consistent I had to quickly hide the other tyres and replace them without being able to notice, the one problem I did have was making it look like a natural fall. Tyres are heavy and fall unusually slow so it took a lot of messing around on the graph editor to make it look like one continuous fall but because the tire has a straight edge I wanted it to look like it was standing straight for a second and then falling again which I think I managed to achieve in my animation. After rendering the animation and adding motion blur to blend the tyre switch better it became obvious what I was trying to attempt and I knew I had to go back and stick to just the one tyre. It took a while to look right as the origin point made the turn of the tyre difficult to control. I spent a long time using the EpicPen to make sure the tyre didn’t move from its pivot point but eventually, I was able to achieve what I wanted. Eventually, I went back to my three tyre method and just lowered the motion blur shutter to make it less obvious.
On week 8, while working on the weight lift the class introduced the jumping animation. In my head, it seemed simple, I figured, how hard can it be? It can be very hard. I found humour in my efforts in class, everything I did looked wrong but what help would being stressed be? My first attempt was shocking, I didn’t understand the body mechanics at all. I took a break and came back, asking Aodhan for help, and I better understood the task. The samples were shown in 2D animation where stretching the body is a lot more acceptable so my legs being stretched didn’t work as well as it would have if the animation had been in 2D but some things are like that. I wanted to do a basketball jump. I made a dumb decision at the beginning of this one, I barely used any reference and just went with a regular jump reference from The Animators Survival Guide and it looked shocking. I didn’t move the character’s body control so he jumped and landed in the same spot which looked really bad. I decided for my jump I wanted to do a simple enough basketball dunk, I talked to Matthew Hamill and he strongly suggested using dunk contest references and also showed me in person how the legs should be and how to convey the power of the jump but there wasn’t a lot of references that the player landed in a regular motion because the dunks were always too extreme for that. Andrew suggested I look at using the C pose while Snow was jumping in the air which definitely added personality to my rig which I was happy with. I sent a copy of my file over to Aodhan for feedback. Straight away he said I needed to fix the landing, my rig was suddenly transporting forward before the land and it looked really offputting. He was dead right, I had been staring at it for so long that I hadn’t even realised. I watched videos of people dunking online but a lot of the landings were way too dramatic to use as my character wasn’t falling down for too long and wasn’t coming from very far away. It took a while for me to get it looking right, I filmed myself jumping from different heights to use that as a reference and that’s when I noticed something else. My arms looked wrong, I tried filming myself with my arms going backwards, like they did in my animation but it felt so unnatural so I changed that to follow my video and it looked so much better. I used EpicPen to draw on top of my animations to make sure the jump was following a curve and edited it when it wasn’t. I understood how important references were throughout this assignment, the difference in my work from when I used references and when I didn’t was blatantly obvious so I made sure that everything I animated had a reason as to why it looked like that. I found the feet difficult to animate, Aodhan told me they were going too far back for a jump that didn’t have that much power so I lowered them and kept replaying, leaving and coming back on a later date to see if it looked good. I think with the Run and Jump animation I was far too comfortable going in. I needed to go back to what I learned in week one and that was to use references as much as possible because without them it wouldn’t look good. I’m glad I learnt it now and not after the deadline, that’s why getting feedback was so important, it made me realise the problem wasn’t that I wasn’t able to do it it was that I had no reference to follow. When re-reading the brief it said ‘Running and jumping over a space or object’, this panicked me slightly. I wasn’t sure if over a ‘space’ meant a space on the ground or a physical hole, by the time I had reread the brief I couldn’t contact the lecturers so I took the problem into my own hands. I could make Rain jump over something or create a massive hole in the ground. Originally I wanted him to jump over another rig but Alec only recommended using one rig at a time so I then went to thinking, what do basketball players jump over? They don’t really jump over anything so I created a big crack in the ground as many basketball courts are falling into disrepair in the area I live in so it really isn’t that unusual I just made it a very dramatic crack.
By week 10 the dreaded facial animation was introduced. I was the most stressed about this animation. Alec and Aodhan constantly reminded the class how important being present for the lectures would be and how much more work would go into lip-syncing. The class started off by downloading a file with preloaded facial poses in the rain rig. I downloaded the sample audio file and began work. I really enjoyed this class task, I felt that all the 10 weeks prior were leading me up to this, the last part to learn when animating a person. It actually took me a while to decide on an audio to use, I originally wanted to take one from a podcast but decided on a clip from Parks and Recreation. I brought the mp3 into blender and went through the list of poses that were premade. I was using references that Alec put up on blackboard of mouth shapes and found some of them weren’t in blender so my first task was to make reusable asset poses and put them into the pose library. I made an L shape, and another W shape because the premade one wasn’t working and I made another A and I shape. After I had made more poses I began to animate. I actually really enjoyed this process and pretty much had the mouth section of the animation complete, I asked Alec for feedback and he told me to move some of the keyframes forward because they didn’t match the audio perfectly. The following week was facial animation with eyebrows and eyes. I didn’t expect this part to be so difficult, everything came down to expression. I asked Alec about some ideas I had and he suggested filming my own reference, I couldn’t use the original video because it was cut off on some parts because it’s a sitcom. I had the idea to go to TikTok, find the audio and see what other people did. This ended up being so helpful, my reference helped for facial movement but I was a bit shy and am not a great actor but people will post anything online so I downloaded some of the tiktoks I found and used a lot of them for different parts of the animation. I really wanted this animation to stand out so I knew I wanted to add as much personality as possible, taking inspiration from the sitcom I made the camera move in fast angles to represent an interview type and to add a bit of humour I made a really bad small version of the Rain rig and added it into her hand to hold like Ben does in the parks and recreation clip for when the rig says ‘would a depressed person make this?’ For feedback I went to my classmates on Discord, it’s nice having feedback from people who are in the same boat and experiencing similar setbacks. I was told by my peers that the eyes were moving too slowly, in class it was explained that eyes dart and move fast instead of slowly panning which was a good reminder.
All in all, I was really pleased with this assignment, it’s bizarre to think that I went from struggling to make a rig walk to animating a full lipsync animation. I’m really proud of the progress that I’ve made and a huge reason for that progress was due to getting feedback from people, it helped me see problems that I couldn’t but also helped me learn new things and new ways to go about animating a certain movement. Another huge reason for the difference in my work was filming my own references, to see how I move in situations compared to references I found online I found taking aspects from both helped a lot more than I thought it would. I won’t be jumping into animation for another while but I definitely would love to learn more, get more advice from guest lecturers and try out some more body mechanic animations in my own time.
References–
weight lift
Run and jump
LipSync