2026-02-06
End it in style, start it with a bang, then fall on your face and win anyway
I've rewritten this a few times, each one invoking a level of irritation that saps way too much energy, so hopefully I can translate that irritation into something productively written this time around in one or two shots, and cut straight to the point. I merged December and January together due to a lot of projects getting carried over into January, and none of them having satisfying conclusions.
Vast majority of this post will focus on the singular thing I've been putting most of my energy into finishing; Aethena.Chimera.Aqua V2. Everything else will come after
Biggest one first, and the one I keep putting off, the "Final" Chimera.Aqua reconstruction. First skin I ever bought for VRChat back on 2025-02-11, and the one that I would rip apart and rebuild over and over to understand how everything worked. When I say very little on this model is 1:1 copied from lackofbinding's version, that's not an exaggeration, nearly everything was rebuilt, remeshed, re-all of it, in part because I wanted to know how to do it all from scratch, given that last year I made thew absolutely reckless move to commit to 3D freelance stuff as a career.
2025 was mostly me working through the growing pains of learning everything from scratch without any direct assistance (in part because, no friends that actually play VRChat, but also I wanted to build up the confidence and willpower to work through tough problems alone without assistance, so I never felt like my skills were being falsely propped up by the direct help of others) along with trying and failing to learn how to pace myself with complex projects. This is why I put "finished" in quotation marks, because I've still not done everything I want to with it, but I need to plant my flag somewhere in the progression of this model, and be open to showing it to others
Here's a link to the thumbnail at full res, along with V1 from early-mid 2025 if curious. I deliberately drew at a resolution that I believed VRChat natively rendered at, so nothing would be downsampled at base resolution. Possibly overkill, but sometimes I also just like drawing at lower resolutions
In the beginning, I only had the idea to edit the textures, but soon I started writing lore in my head about how a robotic diving model might be hijacked by some unknown street artist, and reconfigured to be some kind of vandal, but the end result was clearly not done by a professional. Ripping metal bars off, re-welding them together, getting that dented effect to it, layers of old paint, the identity of someone else scrawled across something mass produced, so it may be a unique bit of self-expression.
In my brainstorming about this lore, way back last year when I was ripping off parts and re-attaching it to other places on the model, I had the idea to rip the dial on the back out, and put it on the left hand, like a watch. By far, this was the most hellish thing to learn and execute on, but I did it, and it's the great big thing about it I love. Talks to VRCOSC, and reports back the time by rotating each hand, just like a real watch. Later on, exceedingly ambitious plans involve a day-date function, and something that maybe looks a little more watch-like, but above all else I wanted it all to look hand-painted, done by an amateur while still not looking like shit, and I think I did that well
Adding onto repurposing a mass produced model to do something it wasn't designed to do, lots of hoses all over the place no longer control oxygen from a tank, but instead control paint from canisters. At some point early on, before I understood the limitations of VRChat, I loved the idea of actually being able to spray paint with the right hand, paint things a certain color like a true vandal, but the idea was scrapped for being too ambitious as someone that at the time, barely had a passing knowledge of how to use Blender.
As someone that's now used Blender for basically a year and some change, all I see everywhere is a bunch of compromises and patchwork fixes to things I didn't know the proper practice of back when I started this project, but hopefully the larger model looks as good as I suspect it should look
When I say I started work on this as a near total beginner, I mean it. I remember the first thing I witnessed that I started to dig to understand the why of, was overlapping UVs to save on space, and what the consensus was exactly. Speaking from my own experiences now and research plus actively working with my absolute hell nightmare UVs that originated from not knowing what I was doing, it's contextual. Overlapping UVs can cause issues later on down the line, but they can also save a lot of UV space and texture size.
The reason I wanted to scrap them entirely had to do with a lot of the models I noticed being very...symmetrical. It made me hyperfixate on working with and thinking around deliberate asymmetry in models, especially if it means standing out from the crowd, or drawing attention through a unique creative idea. Explicitly, I don't *want* overlapping UVs, because I want to bring attention to asymmetry (that TAIL). That and, beginners don't fully know how overlapping UVs can come back to bite them later on if they don't know why they can be problematic in certain cases
I think another problem I fell into in the beginning, was not doing more parts manually, and relying on something like Substance Painter to export and create the texture images I'd create in it, as I didn't fully know or understand what I was looking at. Presently, almost everything in each current texture for this is hand-drawn. Minus AO of course...I'm not that demented, but I have gone back to touch up AO issues I miss on first pass here and there at times.
In the present, there's a lot of different parts about the texturing process that I understand far better now and can avoid having issues with in the future. Like, the edge padding on this is miserable and borderline embarrassing, but best practices for that wasn't something I learned until more recently, and a majority of these textures were created months ago in a workflow I didn't document properly...and eventually, you get attached to how something looks, and you just have to stop yourself before everything spirals into editing and experimenting in an endless cycle that doesn't really get you anywhere
But goddamn, the textures on it really do speak to me on so many levels. The coloring is based off my most recent, and likely to be my longest lasting sona; Cynthia. Orange and cream Florida shark, with the darker color being the top coat, and the lighter one being on the bottom. I've loved sharks in the fandom since my beginnings back in 2017, and even as far back as 2015-2016 when I was just browsing the art, and I can't recall any that just...swapped the darker color to be on the top.
Have I been called a creamsicle shark a few times because I opted to go for a swirl-y texture for the skin on this model ? Yes, and it's worth it. Smudging over and over and over created such a fantastic texture in areas that I didn't want to mess with, where the darker orange looks like it's leaking from certain areas, spilling and mixing with the lighter coat of paint
During the event, I wrote probably close to a triple digit amount of mail, and sent it out into the world, along with hunting down any mail I would see in listings, in the span of a week. This resulted in me receiving 38 pieces of mail from 26 unique save files, which is frankly wild to me that I was able to grab and digitize so much mail in that span of time.
I did not pace myself particularly well on this project, and any screenshots I will show are not the final planned asset, this is just as far as the month took me. Fun things associated with good times shouldn't be worked on under stressful headspaces
Finally, the REALLY cool part, the Luma Glow...
This singular realization of what I could actually achieve with Luma Glow (I don't think I'm supposed to be capitalizing that, but I'm sticking with it) under certain lighting conditions is why I opted to stick to Furality's Aqua shader, instead of burning even more time deconstructing how lackofbindings handled the custom packing of Orel's modular shaders. This is also why there's no Audiolink yet, but there will be eventually.
It was a last minute experimental idea that I wound up burning maybe...two whole days getting just right ? Make no mistake however, it was ABSOLUTELY worth it, and I'd have happily spent more time getting the seams and lighting just right, but again, theme of this post is "Wrap it up, you can't have this looming over you forever"
This is the big hill I will fight and die on, the one that no quarter will be given on, and where there's going to be a massive rant-y block of text. The biggest sentiment I picked up on while learning how these avatars are made, what people prioritize, how they feel, was that optimization didn't really matter, and that the devs were unreasonable terrorists trying to negotiate for their virtual nonexistent hostages for enforcing certain model restrictions.
This is to put it gently, a bad take. Impostors look AWFUL and remove hand gestures which removes sign language, and there exists a strong amount of entitlement from what may or may not be a loud minority with 99th percentile hardware. The only reasonable excuse that exists is not having the time or energy to dedicate to figuring out how to optimize down an avatar to a reasonable degree, and just want to use VRChat to relax. For the average user, fine, I won't argue it, but as someone that wants to do this as a profession ? Incuriosity is unacceptable, and antithetical to self-improvement.
If you held me at gunpoint and for some reason asked me to choose between who to emulate more closely in my work, John Romero or John Carmack, I would choose Carmack without flinching, the end product is a higher priority than the artistic vision. If I cannot optimize it down to Good on all platforms, then it's not good enough to be my work. I don't care how much has to be trimmed off, if we start cutting the fat, the muscle, the bone, if we're hacking off whole fucking limbs and writing lore to explain it, we make it run on all platforms no matter what. That's the standard I want to hold myself to, the sentiment that I WILL get an idea to work, even if it requires me rethinking the entire structure of a project, or learning a few new techniques in the process
To wrap all this up, I'm gonna rattle off a bunch of things that are either bugs, things didn't make the cut on the final V2 model that were either technically on Chimera.Aqua, were side ideas I had to drop, or are things that exist as data I have, but are just not added yet
No audiolink No tail controller No clothing/prop toggle No color control (big priority if I make a cloneable version) Mouth stuck open, but only in avatar preview VRChat SDK claims incorrect download and uncompressed size (uploads fine...) Existing animation data for custom hand poses exists, but didn't add it yet Existing facial expression animation data also exists, not fully implemented Alternative head modification that allows for proper visemes instead of a jaw flap bone Alternative colorations Adding a day/date function on the watch
None of this seems particularly large scale, or even difficult to do, but the problem is how many other threads come loose once you start pulling them, so consider this a little list of what the soonish future will have for this model, given it's basically just...me ? I would like to have some kind of cloneable version in the future for anyone else that finds this model neat, possibly stripped of some of my own self-identifying parts of it, so others can self-insert easier
Okay, 3D over for now so I can comment on a couple 2D things I'm working on in the background, and some feelings about it all.
I've never been a huge anatomy or lineart person, I always loved color theory, shading, and figuring out how to squeeze detail out of stone-shaped canvases. That being said, recently I've been feeling some brutal limitations on what I'm actually capable of in my current style, and I want some change, some improvement. To eat my artistic vegetables, and actually have a concrete foundation for all parts of the process. How I intend to go about this...well, I'll have to figure it out as I go, but I suspect I'll be collecting some new insights to improvement fairly soon.
Oh, I also intend to release my digital brush set I use this year, along with some documentation for getting the most out of it, how I made it, and where I've used each brush in big projects. The complication is that I am stubborn, and still use Affinity Photo 2, so everything's stuck in the ".afbrushes" format, and will require me to learn how other art programs translate standard brush formats. I'm alright with learning that though, I'm not exactly married to Affinity Photo anyhow, and have a laundry list of gripes with it despite now going on 5 years of using it.
Full res versions for Left, Middle, and Right illustrations, if curious
-Pierce.webp){kind=link}
Just kidding, there's more relevant 3D stuff...
Ah, the other giant timesink project. Early on in December, the Poké Classic Network held a Zubat relay event, where a bunch of Zubat were sent out with mail attached to them, because apparently the weird little group I accidentally started had some pretty cool influence. Nobody's bothered to ask me about this since I last shared pictures in December, so I assume I'm the only one who cares anymore, but I still intend to see the next S-S-S update through to the end, even if it does get delayed slightly due to just...a lotta weird medical life stuff happening that kept hitting me right in the kneecap with the starting pistol on a bunch of days.
It's another one of those projects that, while it has progressed past the initial finish line, is still one that has its own spiraling issues I intend to fix SOON, as long as NOBODY ELSE ENDS UP HOSPITALIZED.
*ahem*
Hey, remember how I talked about typing in my last post ? Here's an update on that. I don't mesh well with daily streaks, and have some complicated thoughts about the overuse of streaks being a particularly toxic addition to a lot of services, purely to increase user retention (looking at you, Duolingo, I burned that 900 day streak intentionally, and I'll never regret it for a second)
Combine this with playing a little bit of Tetris here and there, and I must admit that my hands are pretty damn fatigued since last post. I didn't even give myself time to explain a couple little site updates here and there, but...shark's gotta keep swimming.
🧡
- Aethena, 2026-02-06