Creating a Series Pt 5 - Using what I have

Making Of / 09 December 2018


So we're half-way through December and I'm still working on my art-project in UE4.  

One thing I kind of want to get straight is that people look at my work and go: MAN he works fast, he's already got so much going on!!  

Now, I mostly show my progress here and even that I spend time to make sure it looks at least decent before showing it, I don't really want people to believe that I'm constantly pushing out really nice things or live the perfect life or something like that (some kind of weird instagram life).

(Hard-working kitty)

Truth is I spend a ridiculous amount of hours working & researching stuff in my spare-time. Between having a day-job as a AAA-artist at Ubisoft Massive, my mentorship & this I do spend a ridiculous amount of time just working & thinking about 3D right now.

I can't say this enough: Unless you love doing 3D all day everyday , don't do it.

I actually love doing this since my two main passions are:
1. Video-Games
2. Making Video-Games (/Art)

I do this for fun and I'm sure that during next year I will take a bunch of brakes from this but for now, I'm just having fun with it.


The Environment Art Mentorship is still going strong for me as well, I think I have 1 spot left for the January sign-ups. So this week I've been trying to take it a bit more easy spending some time playing video-games and not putting all my spare time into the art-project.

Currently playing Zelda - Breath of the Wild, since I though it'd be good reference for what I'm currently doing. The main surprise of the game I think is that I feel like the Art overall is kinda uneven, but then again the world is super vast and is running fine on a Nintendo Switch with a bunch of awesome cool systematic effects so overall it's a very impressive experience.


So I decided that it was time to use the assets I have to create something substantial. At least I could get started with something. So I decided to just create something interesting in the center of my open-world.  I'm currently sitting on:

  • A couple of rocks & cliff meshes
  • 1 Tree
  • 2 Types of Grass (short / long)
  • 3-4 types of Flowers (Different Colors / Shapes)
  • Blockout shape of some huge-structure I'm working on.

So what can you create with just these? 
I decided that I would be the entrance to an ancient tree-temple that is lying underneath the hill of which the tree is standing on. What you'll see below is basically my first draft of the composition of this area. I spent a 2-3 hours basically setting up this composition and prepped it to be in the center of the world. (which currently is 8x8km)

The thinking behind the composition was that I wanted something to draw the eye-further into the scene in a sweeping motion (Red-Arrows) leading the viewers eye to something of interesting the background (yet, to figure that out properly). In the middle-ground I wanted to visual interests, to the left and right. Left being some kind of adventurer camp set up on one side of the river.

On the other side would be the entrance into the cave leading into the Grasslands Temple. For now I haven't figured out if I want to keep it a cave or if I want to do some kind of temple entrance for it.


I have actually starting nailing down some of the lore for my project as well, though I'm still unsure how I can actually infuse some of it in the designs I'm currently making. I've been interested in kind of merging fantasy with sci-fi, in some way but more often than not it's just leaving me confused as how to meld the two.

So I've also started blocking out a modular kit for a temple , though I have yet to actually decide on an Art-Direction for what I want it to look like which is a major issue. I need to figure out:

1. Temple Style for the Grasslands

2. Village Style for the Grasslands

It's something I need to spend more time on for sure and something that you'll probably get to see some of next week or the week after.


After spending some more time with the initial shot I decided to actually try and upgrade my trees. I didn't want them as noisy or expensive as they were before. I wanted a much simpler shape coming through. So the main difference from this version and the version above is that the canopy is much simpler, cheaper, performant and actually a bit more brightly lit and fluffy looking.

I also decided to try out some of the root meshes that I made for another scene in this one just to see if it would work out and I think that generally it's getting a bit more interesting.

Below is a top view of the area currently, as you can see I've started developing some of the area around this whole place but it also goes to show that I'm trying to think of the whole space. It needs to work as a proper game-environment, not only the 1 single shot for a portfolio piece. I think generally I'm just going to try and develop everything from the center and outwards, coming up with shapes and landscape as I go.

So another thing that is actually happening in the open-world environment is the day & night cycle. 

Just looking at my current result I'm pretty frickin sure where I'd like the sun & lighting to be for that specific angle but when I've got a whole open-world time-of-day setup this becomes much harder to deal with haha.

This usually means that I can't have "perfect" lighting for every location in the whole world, my moon & sun has an actual set trajectory, the only thing I can actually do is to mess around with local post-processes in each location later on when I polish some areas up.

It makes me realize how much more streamlined a out-of-the-box TimeofDay editor for Unreal could be , right now you have to a bunch of weird things with Blueprint to even get something simple running. They probably should have a quick look at like, CryEngine or something like that.


I think my next steps will be to actually look at Temple Modules & Building designs. 

I'd also like to quickly photoshop up some blockout textures for what I would want to work together with the temples & buildings. Nothing complete and they would just be doodles before I actually invest the time to make high-poly stuff or real stuff in Substance Designer.

Until next time!

Warping out!

/Chris Radsby

Creating a Series Pt 4 - Steady Progress

Making Of / 04 December 2018


So December have just started which means I've also started my december Environment Art Mentorship and it's definitely exciting and I'm already super proud of my mentee already. :) 

As for the project , my twin brother Alexander helped me out with logo-design for my art-project.

And I'm quite happy with the direction the logo is going. It'll be well suited for what I'm planning later as well.

I've been making some progress, this week I've spent optimizing all of my current imported assets using the custom-lod workflow for the Grass & Flowers and UE4 Built-In Mesh Optimization Tools for the rest. Having worked professionally with Simplygon before I can easily say that the built-in tools are just as good as Simplygon if not better, which is very impressive in general. 


I also tried to make my terrain material automatically paint itself in the material. I need to make it more advanced since it only paints grass + cliff right now. I probably want to extend it further or even try and figure out how I make sure I get get varied regions in as well.

Here is a video of it

Later on I will probably try and tie some of my vegetation/props to the terrain layers so I get automatic grass placement etc.


The main work going into the development right now is about consistency and getting the systems up and running. Everything I feel is important before I actually sit down and start designing making a specific locations that actually looks good. 

I started researching more about dynamic lighting for the art-style I'm going for and I decided that I should look more towards what's available to do with distance fields. Mesh Distance fields are a 3D volumetric texture (field) that surrounds the mesh that stores the distance from the mesh surface.

It can be used for a bunch of different things and I'm currently using it for:

  • Distance Field AO
  • Distance Field Raytraced-Shadows
  • Distance Field Dynamic Foam on my water.

and I'm looking into more ways of using them in the future.



I've been looking at ways to optimize my world a little bit and since cascaded shadows are quite expensive at a distance as well I thought I'd move over to using Distance Field Raytraced Shadows for the far-distance shadows. Since I already have Distance Fields in, I might as well take advantage of them.

The negatives with Distance Field Ray-Traced Shadows is that they're static, they won't animate with the leaves move etc. So right now I'm using cascade shadow maps for the close-distance shadows and the DF Raytraced ones for the far distance. The quality is worse but from a distance you can't really tell anyway. 


Since I'm working with an actual open world with unique locations/biomes I will have to design I'm currently doing all the prepping I can to make sure everything is working smoothly. I've got a day/night cycle in the game right now as well & I try to see if I can make a consistent look.

Day & Night

Quick panning shot


So I've implemented a quick cliff & rock shader that uses a surface material that scales with the scale of the object itself but also a world-aligned normal map/albedo so that I can use to get erosion lines along the cliffs / rocks if I want too.


The next steps might be more interesting I guess, I'll single out a new location where I will start the actual portfolio piece.

  • Moving forward I'm going to have to look at trees, for real.
  • Deciding on a concept & locations for the new area
  • It's basically going to be my next portfolio piece so there is some serious work that needs to go into it as well
  • Everything from making more vegetation to making more interesting architecture
  • Figure out how to deal with more contrast in the scene without making everything too dark
  • It's still meant to be an actual game environment, complete darkness won't really be an option.

Until next time!

Warping out!


Creating a Series Pt 3 - Early Beginnings

Making Of / 25 November 2018


So people have been wondering about my next project, it's already been started. It actually started 6 years ago when I first made my Animé-inspired piece in Unreal Development Kit. I had gotten hired by Ubisoft Massive and a couple of years later I decided to try and do a stylized environment art piece and it resulted in this. At the time I was heavily inspired by Airborne & Wind Waker HD and I guess even today I still am very inspired by those games.

In reality I have spent about a weeks worth of evenings on this new project setting it up

Right after I finished up Beyond Human, I started researching and gathering material to create another more stylized piece back to the roots of what I enjoyed doing before and then BAM, Helder Pinto shows his Europa-game and it blew  my mind. It's the single-most beautiful game environment I've seen and it reminded me so much why I wanted to make more animé-inspired art.

My main influences:

  • Anime
    • Ghibli
    • Naruto
    • Hunter X Hunter
  • Video Games
    • Zelda : Ocarina of Time
      • Probably the most influential game that made me want to get into games.
    • Zelda : Wind Waker
    • Zelda : Breath of The Wild
    • Rime
    • Firewatch
    • ICO / Shadow of the Colossus
    • Ni No Kuni : Wrath of the White Witch

There is a trend here to the kind of games I really enjoy =P , but yes, these are my main inspirations for the next few months to come.


So I've decided to give the project I'm working on a project-name.
 I've decided to call it Legacy (work in progress logo xD )

Legacy is the project-name for my art-series, it describes a game / world in which the player wanders finding artifacts and legacies from times past.


Well seeing that Helder, has already managed to succesfully re-create Ghibli, I've been analyzing his art a lot and I've been trying to replicate his style in UE4 (I'm sorry buddy :( , but it's just too good, I need to learn your secrets).

So my task right now is to use most of the previous tech that I've used and implement a lot of it in this new Unreal Project but also piece-by-piece trying to figure out how to acquire a look that is either spot-on or similar to Europa & Rime.


So, I still believe I don't have the full toolbox of skills to be able to achieve everything I want. One one hand I'm trying to achieve a new cleaner style and secondly I need to spend more time learning stylized sculpting in Zbrush. So at the same time as I'm doing this in Unreal Engine , I'm also spending time reading / watching / doing tutorials. So you might see some of that here as well.


The long lonely road...

The NR 1 thing I usually do when starting a project is Reference, which I've done and then I'm trying to lay down the ground-work for the whole project, as I stated in my previous post . For me that means, getting all the basic tech in that I want for the scene.

  • Time of Day Solution
  • Cloud System
  • Dynamic Lighting
    • After my recent endevour with LightMaps again for the interior scene I made, I just..want to make art, dynamic lighting helps with that.
  • Global Illumination Solution
    • Ended up looking mostly at Light Propagation Volumes, though Nvidia Voxel GI looks interesting I'm not sure I can be bothered to make my own build to get it going.



I wanted a decent cloud-system this time around so I went shopping around the Marketplace and found
Cloudscape Seasons which I've had great-success with so far, and Nilson Lima has been a great help and support for just setting my stuff up 10/10 service. I'll see how far I can get with this or if I need to change later but for now it gives me really nice fluffy raymarched volumetric clouds.


Color Ramps & Gradient Mapping  has been around for quite some time though I've never really been using it for much. My main thinking for using it in my project is that I want to be more flexible with the colors I use.  I want to:

  • Create Palettes of Colors that I use for everything
    • Grass has one palette, Rocks has one etc etc and each could be tweaked or swapped in other regions of the world
  • Be versatile
    • I don't want to have to push out a specific albedo map out of Substance Painter / Designer every time I want to try new colors
  • Removing Noise
    • One of my visual main pillars are about removing noise, in the environment, limiting the amount of colors any asset can have will help with this overall goal I believe but also help consistency.


The initial implementation of my grass just to use a solid color that I multiplied with a linear gradient, it looked really nice in its simple state. Overall I quite liked the look of it and I was happy with it until I realized it didn't look painterly enough, just samey.

 So after looking at lots of Ghibli references I realized that I need to focus more on the bigger picture  and larger clumps than individual straws of grass.

Now the easiest most obvious solution to getting this type of Shading I realized was to make sure the grass 100% picked up the color from the terrain material. If the terrian material is nice and varied the grass will pick up the colors and be shaded the same way, which gives a nice painterly feel to all the grass.

What's the best thing about this ? It reduces noise but also add more detail to the environment in a good way. The clumps of grass that is colored ends up being in front of another shade of grass which really gives that painterly feel to all the grass.

Another positive about this is that the grass perfectly blends with the distant terrain and it minimizes popping by a lot. It's a lot easier to optimize the grass and make it LOD-away quickly if it already perfectly blends with the terrain underneath.


So I decided to add the water I made for the Beyond Human project though I've stripped out the tessellation since I'm not sure I want to use it since tessellation is generally a real performance hog and I'd like to figure another way making the water generally nicer in the environment. I might want to go back to tessellated water later on but for now just a water-plane will do I think.

I also decided to make some flowers to see if I could get a similar result, I also tried making a quick tree which is still better than I've done before even though its quick. I need to spend more time actually developing the look for some of the trees.


"Imitation is the highest form of flattery"

Well actually, I'm using what he's currently shown as a benchmark of trying to find the same / similar style myself. It's the Ghibli style we want to achieve as well and Europa is the closest piece of art I've found to it. If I nail the arstyle then I can at least move on to my own custom areas. 

 That's the whole point of it! 

Either way I gotta thank Helder for Europas existence, it's inspired me to no end :)


Well, the next step is to basically start developing my own areas and compositions.  That said however I still have to implement proper landscape materials, make nicer trees (current is a 5 min hackjob), have a try at making rocks & figure out shading of buildings and what not.

So in the future you'll end up seeing some more original stuff from me.

Warping Out!

/Chris Radsby