Flowers Scene - February to April 2025

This project grew from a desire to add some more emotional/acting based animation, with 2 characters interacting rather than just one like I had in the digging sequence. I struggled to come up with any ideas for a while, and eventually decided to look at the storyboards for my 4th Year film (that already had a good bit of character interaction, but it's not showreel quality and would take a long time to fix up) and I settled on this scene, and shot new reference for it:

Since I wanted to focus on animation specifically, I used rigged character models from the Blender Studio website. That way I didn't have to worry about modelling or rigging, and I could work with a model that's been rigged to a professional standard. I also didn't intend on modelling any props aside from the bouquet of flowers, and maybe the door but I couldn't resist. I wanted to do a bit of everything. It wound up becoming a lighting exercise and a camera movement exercise as well as an animation one, but both of those wound up being somewhat limited since I didn't plan for them from the beginning. I'm not entirely sure how I feel about it overall. I'm happy with the animation but I think the rest feels a little overdone and a little underdone all at once. And I sort of wish I'd given more attention to the wider plot/gag instead of focusing on making a few good reaction shots.

The Bones and The Beans - September to December 2024

Digging Sequence - May to July 2024

After doing a few walk cycles and simple animation tests using public domain models from the Blender website, I started to feel like I could get more out of it if I was using a character that I had designed myself and that I had some sort of personality in mind for. The most developed one I had was the Cave Spider fella, so I drew up a reference fairly quickly and got to modelling. Somehow, despite being the first full body character model I'd done since midway through 4th year, it went fairly smoothly. The face was a little difficult to figure out and even now I'm sure I could've done a better job, and figuring out the correct neutral position for the arms took a little trial and error. Beyond that I was surprised at how easy I found it.

Next was the rigging. I'd cut a few corners when working on my 4th year film and certain parts of the rig were a little dodgy so this time I found a good tutorial series on YouTube and that made things a lot clearer to me. I wouldn't say I could do a flawless rig from memory right now, but I think I'd get a lot closer than before. I hadn't really intended to do IK for the hands but I'm really glad I did, and I probably wouldn't have thought of it at all if I'd went ahead working from memory or if I'd used a simpler tutorial. Even with the tutorial, though, gimbal lock was still a bit of an issue. I'm not sure if I had the bone angles aligned wrong or if I was using the wrong Euler but the hands in particular didn't work as I'd have liked (rotating the hand to the left or right as if it was waving never really worked if the other axes had been posed). I definitely could've played around with testing the rig a bit more because I didn't realise there were problems with the hands until I was already animating.

As for the animating, I started off with some simple enough stuff, like a walk cycle. I don't think I used any reference for this. I tried but I got too bogged down in copying the details, and the timing didn't flow as evenly as I'd have liked so I just went with the Richard Williams timing and made alterations from there. It was the same story with the run. I did the horror movie spider walk from reference though, and I also struggled a little with it to give it even timing that felt natural for a cycle. I certainly could've spent more time polishing it.

After doing those tests, I had a look at the rough storyboards I'd drawn up for the Cave Spider project to see which shots had interesting acting. First I tried doing a shot where the character sat down dejectedly, but made a silly mistake in using reference that was shot at 30fps, which didn't flow properly when played back at 25. I gave up on it then and started working on another shot where he was digging a hole. A lot of the shots I'd planned didn't involve all that much acting beyond walking, and some of the ones that did were more face based and I wasn't confident that the rig would be capable of doing that effectively. The digging shot gave me an opportunity to try and practice animating different arm movements, and using an IK arm rig.

Working with arms that were oddly proportioned presented problems, especially since I'd shot reference with my own normal length arms. Most of the time I tried to match it to the reference as closely as possible but thinking back I should've based it more on how arms like that would likely move.

I'd never really properly messed around with the graph editor before doing these shots. I'd never had the time to go beyond setting key frames on any other project I'd done. Looking back at earlier tests it certainly makes a huge difference. I'm sure there's more to learn about it but I feel like I've got the hang of it and I'm glad to have finally tried it out at this embarrassingly late stage.

Once I'd done the animation, I moved on to the texturing and lookdev. The textures are all procedural noise/voronoi based textures made with Blender's shader nodes, with the exception of the bump data subtle footprints in the sand that I made in Clip Studio. I had a rough idea in my head of what I wanted these shots to look like, and some concept art from a couple years ago to work off but I mostly winged it, which probably isn't the best idea generally but I think it worked out fine here. I messed around with particle effects a good bit just for the craic. The grass could do with a little work still, there's some weirdly shaped blades. I'm not sure how important it is to know anything about particles but I figured it wouldn't do any harm to experiment with them a bit after finishing everything else