A peek into Drakkenridge Dungeon design and tech.

A peek into Drakkenridge Dungeon design and tech.

I've regularly posted updates giving a sneak-peek in my process for the creation of VR mobile-friendly assets on Twitter/X (@luxregina) or more recently on BlueSky (luxregina.bsky.social) but the short format isn't the best for deeper dives... So, let's inaugurate LinkedIn articles shall we?

DrakkenRidge?

Since I've never really posted here, let's quickly recap: after shipping Stones of Harlath (2021) and Shock Troops (2022), I've started working on the next project (with Simeon Acker ).

DrakkenRidge is an ambitious VR RPG that features loads of different locations, with a low-poly/pixel art graphic treatment. As we get closer to release, I'll post more about the game itself and its design.

You can check some of the other environments on my ArtStation > artstation.com/luxregina

In the meantime, quick gameplay video here > https://youtu.be/esEdOCGAjm0?si=7VF1DGb0ReWnp9t1

Setting up the mood

At this point, the Dwarven Forge being the penultimate dungeon, I have a whole bunch of assets that can be repurposed, allowing me to block out my dungeon relatively quickly.

Each dungeon in Drakkenridge has a mood, a particular color palette, and it's own "dramatic" lighting. I wanted the forge to leverage tones of yellows and reds, and I knew a lot of that would be due to the lighting scenario caused by lava and fire, a light coming from below, so I decided to experiment with lighting scenarios on a previously completed dungeon, a dwarven mine, happening to share some assets with the Forge.

This is not something I usually do, but as mentioned before, because how late it is in production, I have now a lot of assets handy to prototype my levels before actually modeling the proper level assets.

Above: the Dwarven Mine dungeon, Below: the same, but with the Dwarven Forge lighting scenario

Building the dungeon

At this point, I'm starting to have a very good idea of the look I want for that new dungeon and start working on the low-poly models. Not much to say, I can write more about the modeling aspect if needed but it's pretty straight-forward, aside the Atlasing process which I will cover later in this article.

As I create the assets, I bring them in Unity and quickly mock them into a concept environment (something that won't survive in the game, and is not implemented in prefabs, etc) just to check how all the assets are working together, the cool stuff I can make by combining them, what's missing, etc...

A visual "player Zoo" of some sort, meant to test how the various assets can be combined together.

Once I feel I have most of the needed assets, I then start implementing them more correctly this time, breaking them intro prefabs in Unity, creating groups, etc.

I also use this sequence to refine some of my lighting. At this point, assets are not UVed yet, though...

Starting to build the actual dungeon, with a secondary lighting pass.

Atlasing the UVs

This is the part people are often curious about. I've posted about it in threads on X, and even made a short video showing the process > https://youtu.be/G1BsghZD-08?si=XzpR-S2z2Hs7W9Tx

The question often are "How big are your Atlases?" and "What if you cannot fit everything into one Atlas?"

For how big they are, well it depends on all the atlases together, the variance of assets sizes, the texel density that often needs to be precisely preserved because of the pixel art (disparity in texel density would be VERY visible with my pixelated textures).

Atlas#1 combines all the "small" items into one 2K atlas.

If all assets cannot fit in one Atlas, then I create another :) In short, a level can be composed of several atlases. For now, the Dwarven Forge is composed of 5 Atlases, most being 2K resolution, except for one that holds very large assets that is 4K (again, texel/pixel density needs to match).

Color-coded atlases in the level (the 5th one isn't shown because it's not implemented yet)

In the example below, you have 3 atlases present: n, n+1 and n+2... the density between n and n+1 is identical... n+2 atlases big items, like large rocks, etc... and I deemed it OK that the density would be 2 to 1 (so the pixels on the rocks are larger)

Finishing and polishing

This part is pretty straight-forward (at least, to me!) I can write more about it if this article resonates with people :) In short, the rest of the time is spent painting the actual pixel-art textures, while set-dressing for tertiary level of detail, while adjusting the lighting, then correcting the textures, etc - rinse and repeat!

I'll leave you guys with some "work-in-progress" screenshots showing some parts of the dungeon I've been working on the last couple days.

Cheers!




Anna Percy

Game Development Major at Savannah College of Art and Design. | Concept Art and Technical Direction minor.

3 个月

This is absolutely inspiring to go through and read!

要查看或添加评论,请登录

社区洞察

其他会员也浏览了