Computer Arts Practice

3rd Year Uni Module: creating a cute 2D bug character for a platformer

[pxlbcap-blog]

Final submission~

[pxlbcap-blog]

I haven’t been posting the spritesheets until now because they’re not as clean to view as animated gifs I don’t think, as the gifs can get across more immediately what it looks like. However, the sprites are there and they are required to be handed in for the brief, so just posting them here as well for the sake of including everything in the blog, too.

[pxlbcap-blog]

So I made some changes to the dash after getting feedback from my lecturer. I extended the whooshy lines as he suggested, to try and enhance the feeling of speed. I also added a lot more movement to the mantis (lifting him off the ground slightly and pushing him forward) as well as stabilising his landing and step back, which was a lot wonkier before, and changed the eyes to more appropriately suit what he’s doing, and smooth them up a bit.

I also got the dance animation finished. Any amount of research shows that real life mantis’ “dancing” invariably involves the scythes being raised up excitedly and the wings spread. I also wanted to maintain a nod to the original influence of dancing spiders who move and lift their legs. I thought this combination would be a good way to get across the feeling of dancing and light heartedness, without actually taking too long for the animation itself. I wanted to keep it nippy, keeping in mind that it’s designed as a sprite for a game which a player would be controlling. I also thought that the legs raising one after the other is something that could actually be programmed in to player controls, meaning the player would need to manually alternate buttons to raise the legs and be a bit more involved than simply a contextual button press.

I also wanted to make this the one animation where the mantis just looks happy and like he’s having fun. The character design was to create an endearing and emotionally expressive character, but so far he hasn’t expressed anything other than cute wariness and cute determination. I didn’t want to risk leaving out cute happiness, so tried to make him just look smiley and innocent!

[pxlbcap-blog]

So, this is an attempt to neaten up the previous mess by working After Effects instead. I think it came out looking a hell of a lot smoother, but will take some time to get the rest of the cycles put onto it properly. Again, that can be a tomorrow problem though, as I still need to actually get the Dance cycle finished and want to try neatening up the sprint at least.

I’ll see what I can do before coming back to this any more. I mainly just wanted to test this out on AE while I was on the software for another module anyways~

[pxlbcap-blog]

The cyclicals so far, with some touchups across them all since I’ve last posted them. Nothing major really, I’ve just noticed some wee things like a lack of smoothness in some leg movements and eyelids and such, so have changed single frames here and there to clean it up.

With the jump I’ve tried to imagine it more as a dash skill for the player to just rattle quickly as an escape method. I went with a blurry whooshiness to add some congruity with the spin whooshliness but think I might touch it up as it looks a bit too splashy, I think.

[pxlbcap-blog]

So after putting this together more fully with 4/5 of the cyclicals, I’ve decided that if I want to display a backgrounded showreel this way, I should probably use different software, as this didn’t work out that well. I think the cyclicals look better on a white screen than the way I’ve presented them here, and I feel like I’ve lost a chunk of the character impact they had, just because tey’re not settling quite as nicely.

Well, to be fair I think I could neaten it up a lot from this as well, but am rushing a bit because I don’t want to devote too much time to something that isn’t essential and just there for prettification.

Also, there’s no shadow under the character which I can’t help but noticing, but the software I use for pixel art doesn’t handle translucency for shadows and I didn’t want to stencil in shadows until the movements were completed anyway.

I’ll look at recompiling this in different software over the week/end to see if I can come up wth a better result than this, but I am going to prioritise finishing the last cyclical first.

[pxlbcap-blog]

Right, so decided to start trying to see about compiling the cyclicals onto the background towards the final image. Unfortunately, Sketchbook doesn’t support 15fps which is what’s required in the brief. So, I’ve taken the frames (which I had on twos already) and doubled them up to fours while playing on 30fps so it can maintain the correct look. Not ideal when moving every individual frame.

Anyway, it was going generally well, things were looking good in Sketchbook, but upon actually exporting the test file of a few walk loops, I realised that the GIF file was compressed tae fuck, and the video files don’t look quite as sharp as I’d like them to for pixel art.

In short, I’m not sure what I’m going to do yet. I’ll look into other software I can use, and work out if I can use After Effects or Premiere on the Uni computers.

[pxlbcap-blog]

Wanted to create more of a planimatic for the showreel of cyclicals. Most of the animations are done, but for the sake of thoroughness I wanted to add this down the now to make sure the plan feels right at this stage.

[pxlbcap-blog]

So rather than entirely making the showreel of compiled cyclicals plainly on a white background, I thought I’d create some background art for the level itself to represent a potential level for the character. This should also help to solidify the pixel art feel and atmosphere, as well as aid creating the style of Rayman.

Hopefully!

[pxlbcap-blog]

So, after a fair amount of deliberation, I’ve decided to try out what a somersault-esque roll might be like, as it should fit far better the idea of the dancing mantis. Perhaps he’s just showing off, rather than actually rolling through a small hole? Maybe he’s just that badass, since no other mantis can do it?

I went with a blurry whooshliness like comic book style as I felt that it would help translate what’s happening to the mantis easier than trying to actually place everything in a way that looked realistic, when real mantis’ don’t actually do forward rolls. I didn’t feel it would settle properly with the player (uncanny valley) unless I added an appropriate amount of cartoonishness to it to make it just “feel” better.

I also felt that this would nicely fit into Rayman’s cartoonish and exaggerated style without standing out too drastically.

[pxlbcap-blog]

Realised I hadn’t posted my planimatics so far, as they’re expected deliverables that I haven’t actually mentioned anything about so far.

The walk cycle and jump have come across as intended really. The dance has gone through a few iterations (including the animation posted as the high resolution mantis from a while back) but remains mostly focussed around the wings flapping and the mantis rocking back and forward. This suits both how mantis’ in general move (their see-saw-y motion) an is akin to the spiders dance, which is very stationary on the ground, but shakey and movey.

The forward roll is what I’m having the most trouble with, as it’s not really something that mantis’ do, or suits their body type. I’m aware that this is my choice/fault, but it has created a problem nonetheless. I’ve considered having him roll forward and spin into somewhat of a side roll, as well as essentially just jumping and doing a wee somersault. I’m not sure yet what to do for a final version, and might have to rely on advice from my lecturer before advancing too far in that regard.

[pxlbcap-blog]

For the jump animation, I wanted to give the mantis a sense of some determination. I’m going with the notion that our mantis is quite young and friendly, and even the notion of jumping across a stream could be more effort for him than for a grown adult, so I wanted to try and convey that sense of effort....in as cute a way as possible! I also wanted to avoid falling into the trap of just making him look young, lost and helpless in every animation in an effort to make him look cute, and thought that some determined eyes might help give the character a bit more depth.

I’ve also gone in and put the animation on 2s, to keep it the same speed but running at 15fps. This was also a good opportunity for me to go in and smooth certain things out, such as the eyes opening and closing, or the scythes moving into position a bit less jerkily.

[pxlbcap-blog]

After several iterations, I’ve finally come up with a walk cycle I fairly like. I think this one captures fairly well the notion that my mantis is quite timid and wary (since he dances instead of fights) which helps aid the requirement that he be cute and endearing. I’ve added some eye flicks to help push the idea that he’s keeping a careful eye out everywhere he goes.

[pxlbcap-blog]

I had a talk with my lecturer today regarding whether he agreed with, or was at least happy with the new version of the Mantis. I explained that I am quite keen on getting the character concept absolutely nailed, as pixel art is a tricky thing to advocate for. I didn’t want the character concept looking flawed, and risk it coming across as I just went for pixel art as a lazy version of animating. I wanted it to be clear that the pixel art has a purpose in getting across the character in a way that regular digital art doesn’t, and if the character design doesn’t give that, then no amount of animation will help hide it.

Anyhoo, I was given the greenlit to go ahead with animating! I told him that I’d already actually started on a test/concept animation of the Mantis, but just didn’t like it and gave up on it, with plans to start from scratch.

He told me to upload it onto the blog anyway.

I was aiming for a more jerky and jumpy look than the original lowres version had (which was more of a smooth and natural walk) since after watching a lot of videos of mantis’ walking, I realised that they sort of just see saw their body and limbs a bit. They can move very quickly and snappily, but that’s usually when they attack or run, rather than just their casual uptheroad walk. It also doesn’t help that it seems to suck all the cute harmlessness out of it, and give more of a realistic/natural or creepy-bug vibe. I don’t think it worked though, although the current animation isnt what I’d call finished anyway.

[pxlbcap-blog]

Self-Evaluation Strategy!.....At Last!.....Is it?

Right so, weeks ago we got the assignment to create a SES which I totally forgot about. Since it’s Feedback Week, I’ve been reminded that I should really have it done by now, so that’s why it’s this high up on the blog.

Anyway, I’m not entirely sure if this is what’s expected, as I think it might need to be a much more general SES than I’ve proposed here. I’ve decided to do one specifically for Pixel Art partly because it’s I can’t quite help being retrospective since it’s already super late and partly because it’s the same lines of principles I’d be applying to my informal SES anyways.

Also, considering that pixel art is the primary focus of this module and a couple of other projects, I thought it would be a good way for me to display my SES in a practical and relevant way for my current coursework, rather than writing one that might feel uselessly generic. 

I look at things in in the same sort of way I’d look at animation in general: beauty passes and inbetweens: get the basic and general outlines right, and gradually working the more specific or required sections in as needed and only restarting or adding extra passes above it once I’m satisfied that the current pass is as good as can be or that it simply can’t be become what I want it to, and a restart is required.

[pxlbcap-blog]

Back on point, I’ve been looking at trying to get the point across as equally as before, but with a lower pixel density. Good pixel art is about giving as much detail and information about the art through either the suggestion of the image or by visual cues. I felt that, although the previous high pixel density concept was pretty good at getting across the idea of the mantis, that the pixel count was simply higher than it needed to be. I also felt it crept slightly too closely to the pixel threshold whereupon it starts to look simply like pixelatted art, rather than pixel art since the movement of single pixels stopped really having an effect on the full image.

I tried making another version based exactly on the previous concept, but at roughly half the resolution, to see if I could get the same basic idea.

I feel that this is about what I want to work with, as the basic mantis concept is displayed decently, the same colour palettes can be used and the same extra details (such as scythe tips and skin highlights) are maintained, but at no extra necessity for higher pixel resolution.

I've also found that the decreased space has forced me to work more on the chibi characteristics and focus on keeping the character looking cute and harmless, whereas I've previously been overly free all the spare space, and feel that it had snuck up on becoming less chibi and more "real".

I’ll go over it with my lecturer during the week to see if he agrees with my plan, but I’m thinking this seems like the ideal amount to work with. Assuming my lecturer is okay with this plan as well, I can begin animating more seriously from herein. 

[pxlbcap-blog]

The art test assignment (for character concept design) was to take a sketch created by Sergey Kosolov and take it forward into a character concept. This assignment isn’t graded and we’ve been told (perhaps unofficially?) that we don’t strictly need to do these, but I wanted to take part in at least one as it seems like a really useful practice tool for something that was total news to me. I'm not sure if this is sufficient detail for the submission, but I've had a super hectic week and only got around to checking this on the last day before submission. I've tried to do it quickly in the space of a couple of hours before bed, since I’ve got work early in the morning. Considering that, I also basically just traced over the original sketch to quickly get sizes and proportions and hope this is acceptable.

Either way though, I wanted to submit something even if it is unfinished and without colour.

I do intend to give a couple of the other art tests a crack when I get some time!