Lilly Devon did a very detailed overview of the way she creates beautiful little environments in UE4.
Lilly Devon did a very detailed overview of the way she creates beautiful little environments in UE4.
Introduction
My name is Lilly Devon and I’m the founder and Creative Director of a tiny start-up video game company called Little Wolf Studio Ltd. I’ve been in the games industry getting on for 12 years, with a background in game design, art management and production. I’ve worked on many games over the years, from AAA single player console titles to MMORPG’s, mobile and iOS.
My passion when it comes to games, are the emotionally driven stories set in evocative and exotic lands, which completely spirit you away. I’m an extremely visually motivated person, so I always find I’m drawn to games which have a strong art style but with a message behind the experience. The Journey’s, Abzu’s and RiME’s of the gaming scene.
We are currently developing a small demo/showcase to present to a select handful of friends to begin the journey of a new AAA indie game adventure.
UE4 Projects
Recently I started off a small series of what I call my ‘1 Day UE4 Projects’. Initially it started off as a surprise for a concept art friend – this was essentially me trying to recreate a 3D version of one of her wonderful scenes. I gave myself 1 working day to complete it (this was spread over several days in small slots to fit around my normal work) and as I’m not a 3D artist, I had to use whatever objects/VFX/primitives I had to hand. Mostly this was assets I’d previously bought from the marketplace, particles I’d made from scratch and basic objects inside Unreal.
I had so much fun with my initial project, as it was such a tiny self contained experiment. I had no idea if it was going to work, if my friend was going to like it or even if I would be able to manage it within a day. But once I started, everything started coming together and the reaction to the finished piece was amazingly positive.
It started me thinking that perhaps I should do more of them and turn this into a learning experience. I wanted to not only show myself that I can make pretty things (within a very small timeframe) but also show others that they can do the same.
I only started learning Unreal properly this year (as my background was mostly management and paper/written design) – so this was also proof that with a bit of commitment anyone can learn to make interesting scenes using Unreal or similar, pretty quickly.
So, after the first project I started thinking about what kind of things I would like to make. Each mini-project has one element that I have to stick too – whether that’s a certain vibe, a piece of reference or a colour palette – I always want to be learning something new in each iteration as there’s SO much you can do with the right knowhow and tools.
One of my main passions is colour and lighting; it can completely change a scene from something bland to something magical in a matter of minutes with a good initial setup. So I knew going forward that each project would need quite a bit of my time focused on these elements.
I also wanted each scene to tell a story; this could be one I put before the viewer or one that is hinting at something interesting. Every image needed to evoke a feeling, mood or tale; this was one of my main goals. I felt if I could take a simple concept and add some subtle intrigue and broad dynamics I could improve it tenfold. With the right approach, even the most mundane and uninteresting things can tell a beautiful story.
Production
My process and approach to each project is fairly varied and unique to each idea. For example, my first project was working from a piece of concept art – so I had everything there that I needed to follow. This is really the simplest approach for anyone starting out. The challenge however, was to translate that into a 3D scene using Unreal’s landscape editor and trying to work out how I could build the scene elements as well as using tricks that would save time and cover for my lack of 3D modelling experience. The use of silhouettes and alpha planes were important to give the appearance of more than what was really there.
My second scene was inspired by an image from the new Blade Runner 2049 film. This was slightly trickier, as I was trying to evoke a certain vibe and colour palette. I knew that I wanted to include water, neon, plenty of foliage and make an enclosed area – so mostly it was playing around in the space until I had something that felt right and would give an interesting stage for the neon lights which were an important feature to evoke from the reference.
Some of my other projects (like the scene with the trolleys) I roughly sketched out a layout on paper first. Sometimes having full creative freedom can dry up your inspiration (you’d think it was the other way around!) so – having something to concentrate on as a starting point really helps tunnel things in the right direction. The creative process is often an evolution as you build the level – you realise there’s more potential for creating a story. Initially I started that level with the idea of including billowing red and white chimneys instead of trolleys – however I felt it didn’t quite work and the trolleys came after.
Here’s a quick snapshot of the progress of that scene:
Every scene needs a focal point. An X if you will. That can be an actual object or perhaps an interesting vibe/colour pallet. For example, you look at any well established concept artist and study their pieces – you’ll find that even the speed-paints have something fabulous going on. Colour placement is there to draw your eye around the scene, depth and focus for a sense of scale and then the actual elements inside the space. It all adds up to telling a story visually – and that is your X. Once you know what that is (and this is obviously personal to each image you make) then you can build your level around it. Using fundamental composition techniques is key to creating a dynamic and interesting scene, but sometimes it’s better to break the rules and experiment.
Inspiration can come from anywhere. I take elements from films, other games, books, concept art or sometimes just images I have in my head or even something I’ve dreamt. Never be restricted by your medium, as you can create whatever you want with the vast amount of tools freely available. There’s a massive amount of knowledge out there. If I ever get stuck on something, usually there’s a YouTube video to help or an article to point you in the right direction.
Assets
Nearly everything in my projects are from the UE4 marketplace. It would be almost impossible to make something like this in a day from scratch. Part of the point of this for me, is to show how easy it is to be creative and to just have fun with what’s already out there. Obviously, you need a bit of an understanding of how a level editor works in order to set everything up and make it all ‘fit’ – but apart from that it’s pretty simple. You can make something interesting with just the basic primitives within Unreal itself, with the right lighting and mood. That’s really the beauty of it!
The items I have bought normally get pulled around a bit (texture swapping, re-colouring, resizing, scaled and possibly used in ways not intended) – I like to push things to the extreme first and then scale back. Things like particles and VFX I normally make myself – but again some are bought. It depends on how much time I have left. Again, mostly all of this is done inside Unreal (using blueprints/shaders) and hardly ever taken to any external program. The only exception to this is Photoshop, which I used to create some sprites and particles and clean-up/4K scaling for the final image.
Colors
I’m an absolutely huge fan or colour and honestly it can make or break any kind of medium for me. I have an extremely critical eye of how colour is used in film, games, art – you name it I probably have an opinion on it! I find colour theory fascinating, so often most of my time is spent say; deciding on what shade of pink I should make a blade of grass. If colour scares you or you’re not really comfortable with tones, complimentary colours, shades, contrasts etc – then work with a colour wheel. Colour is what ties scenes together and if it’s not right, can throw the whole look off. It’s what draws the eye, what often gives off a mood/feeling and there’s so many ways it can impact your work. You think of games like Inside or Limbo – would it really be the same if they’d of used garish primary colours? Sometimes using strict and muted colour pallets are key in telling your story.
Generally I will change each asset to fit my colour scheme as I’m working as it’s quite an iterative process. Sometimes changing one thing will mess up another, so I find it best to do this as you’re working. The main pass comes in at the end when I apply post processing and LUTS (Look Up Tables)
For these projects I normally only apply one post-processing box around the entire scene and then change the values inside. My main hack for getting these scenes to really pop is to use LUTS. You can make these yourself or buy them in packs from the UE4 store. Each LUT can be edited and changed, so you really do have complete visual control over the look and feel of each space. It doesn’t hurt to use Photoshop to enhance the image contrast and manipulate the colour values or maybe add a vignette if you’re running out of time (however all of this is achievable inside UE4)
Here’s some examples of my most recent scene with the LUT values changed.
And here’s what I ended up with:
Light
In general, my lighting is actually pretty simple compared to say a scene that has taken a few months to make. There are a few reasons for this:
Since my scenes are for the purpose of a singular screenshot, I can get away with only using a few key lights in certain areas. I also use spot lights and point lights to pick out parts of the scene and ensure that nothing gets lost. In theory, you can get away with just using one light source. The main reason for me using smaller lights in my scenes is to highlight certain areas or objects and to bring extra colour as well as highlighting an interesting shape, like the foliage in the Neon room. In the trolley scene, there’s lots of point lights to enhance the ambience and to generate the natural behaviour of reflected light to bounce more interesting colours around from the water. Don’t forget to use Planar reflections or reflection spheres/boxes in times you’re in need of an extra visual pop also.
Some lights are actually Light Functions (these are used to project textures from the light-source and are used to generate caustics and the fan shadow in the Neon scene. I always colour each source as I never use white light.
You always have to build your lighting on each iteration of the lighting setup. You can’t wing it; it has to be done regularly, as you need to know how things look once built. I work on a very slow computer, so even with my build times on Test – it can take up to 2 hours (especially with a landscape full of alpha grass). Make sure build times are considered when planning your project.
The amount of lights/type of lights I use differs for each project. I will usually start with a sky light and directional light. My Directional Light colour is normally set to a soft off-orange by habit as I always find it looks so much better than a straight white tone. But this is just personal preference. For my Blade Runner scene this wasn’t obviously the case as it was actually a super saturated Barbie pink.
For the Point and Spot Lights, I position these in areas I want to highlight. So – places I want the viewer to perhaps look, or areas I felt need the light or colour lifting slightly. Adding post processing has a habit of dulling light sources and squishing colours, so most of my lights are at their highest luminosity. It doesn’t really make a huge amount of difference at what quality you have the light source set at – but most of mine are set to dynamic to allow nice shadows, fog and particle movement (mainly just in case I want to make a Gif)
Time Costs
The biggest challenge for making these 1 day projects is bang for buck. Essentially, what smoke and mirrors I can get away with to make the scene pop in the least amount of time. I like the fact that I only allow myself a very small amount of time and seeing what I can achieve within that. My everyday weakness is fiddling around in a scene until it’s unrecognisable from where I started out – sometimes that’s not a bad thing, but with this I literally just have to STOP and pens down. A challenging task for the majority of us! When I come across issues, I take to the internet to search down possible solutions. Sometimes I just can’t make things work and I start again. There’s no shame in tearing everything up and starting from scratch. If anything – this is supposed to be liberating and making what *you* want, so if you don’t like it, bin it. We are often our own worse critics, so go with your gut.
I really do encourage anyone to have a go at this. Whether you’re a seasoned game developer, a graduate, a kid still in school or just someone interested in how this all works. Your only limit I guess is your wallet when buying stuff from the Marketplace!
Think about what inspires you, what drives your passion or how you can challenge yourself. What ideas and crazy things would you make if you could? The bottom line is – you can use level editors like Unreal as an extension of your creativity. You don’t have to be a 3D artist to make something interesting. What stories do you want to tell? Play with colours and lighting – don’t be afraid of doing something weird! You’re never too young or old to try something new. Experiment, be bold, be brave and most importantly, have fun!