João Kouto has shared his experience in lighting art, told about taking CGMA's Art of Lighting for Games course, and revealed his approach to assets and lighting in his projects.
My name is Kouto, I’m from Braga, Portugal. My passion for videogames, cinematography, storytelling and world-building is leading me to pursue a career as a lighting artist. I took a Graphic Design course at IPCA university, for some time, I was a freelance artist and also worked with a small studio as a graphic designer. I slowly realized all of this wasn’t what I really wanted, so I got really into concept, environment, and lighting art during my third year in university, and after that, I took the CGMA Art of Lighting for Games course in order to improve my artistic and technical abilities and to help achieve my objectives.
CGMA's Art of Lighting for Games Course
Since my focus is to get into the videogame industry as a lighting artist, I went out and tried to figure out what other junior lighting artists did portfolio-wise in order to succeed. I noticed a lot of them did relight of existing maps from the Unreal Engine marketplace and also did multiple lighting scenarios. Particularly Mara Cerreti drew my attention because some of her work was done during this exact CGMA course. I got really curious about it and doing some research I realized that it could be a great way for me to learn and improve as an artist.
The course teaches us a lot about the basics of lighting and how to set things up as well as explains a lot about the technical part of Unreal Engine 4 for us to get started. There is a big emphasis and focus on the artistic side rather than the technical side of lighting. The focus is a lot more on art, player navigation, composition, color, tones, and a big emphasis on how important it is to do research and get references before doing anything. It is much harder to teach art than the technical side of it, as my great mentor Peter Tran would say.
I think the best part of the course is the mentorship. Having someone every week reviewing your work and pointing out what the good, the bad and the ugly is and how you can push it to the next level is crucial for every artist, and that definitely helped me improve. Not only that but the fact that we have a Q&A every week for about an hour with our mentor where we can ask any sort of question regarding lighting, art, job interviews, portfolio, or anything is really awesome. Information is everything and the more questions you ask the better.
I do highly recommend one of these courses, mostly because I firmly believe mentorship is crucial for all artists who want to improve and get better at any craftsmanship. But keep in mind that at the end of the day, it’s all on you.
Inspiration and Goals
My goal with the "Lighting, composition and set dressing studies" series is to create a one-shot composition that would let me explore a number of different lighting scenarios. I also tend to do some set dressing work on these lighting variations if I see that it makes sense to do so to help improve the overall mood I want for the scene.
A lot of my initial ideas come from writing. I love to write my own short stories or just a bunch of thoughts or feelings that I am having just about anything. It helps me alleviate and make more sense of all the voices we constantly have in our head and put things a little more in order. I also like to make some very simple and quick composition sketches in my notebook. They don't necessarily even need to be about the specific scene I'm making, but it's just some very quick and loose exercises.
I also end up getting inspired by a lot of different sources that range from concept art, movies, videogames, graphic design, and photography the most.
When I have a certain idea of what I want to do, I make a mood board with different images to help me better understand composition, color, tone, lighting, or anything else that I feel it might help in some way.
Direct visual references for The Park scene
Apart from these more direct references, I have been building a library of multiple references that I can easily pick up. And I absolutely recommend everyone to build libraries for literally everything that they find relevant for their work. There is a ton of websites and tools you can use to build those. I have a lot of collections on Artstation, Shotdeck (an incredible website with a ton of information about specific cinematic shots), and even Instagram which is also an excellent source of inspiration for photography and graphic design.
I would like to draw attention to the world of graphic design. There are a ton of graphic designers out there that do some incredible work. By analyzing their projects, we can learn a lot about composition, color, and especially how to tell a story in a quick and effective way discovering the techniques that they use.
There are a lot of ways to tell a visual story, using color, shapes, lighting, negative space, composition, etc. Keeping all of these in mind is super useful and insightful.
In all of my projects, I use Megascans, Sketchfab, or Unreal Engine marketplace content. Using all of these amazing assets lets me build these images a lot quicker and with a lot more iteration. Although I have learned the basics of asset-creation, it’s just not something that I want to focus on and something that aspiring lighting artists should keep in mind.
You don’t need to do any sort of asset creation for your lighting portfolio. It’s just not something that companies are going to look for or really care about when looking at it. I focus on set dressing, composition, and lighting because all of these elements go hand in hand with each other. I like to think that I have an approach to my work the same way as a cinematographer has to movies and I do want to improve on all of these abilities to improve my visual storytelling abilities.
Assembling the Scenes
As for composition, movies are my main inspiration here. I absolutely adore the visual work done by Akira Kurosawa with movies like Throne of Blood, Ran or Kagemusha, or other directors like Sergio Leone, Andrei Tarkovsky, and many others. I analyse a lot of these old movies since they didn’t have the cameras or editing abilities that we have today and they focused a lot more on delivering interesting compositional work. Not to say that today we don’t have amazing directors and cinematographers, because we absolutely do, people like Roger Deakins or Greg Freiser, for example, are incredible at what they do.
Working with Unreal Engine's Cine Camera Actor gives artists the ability to manipulate the aspect ratio, focal length, aperture, and more. Learning these basic elements of photography can really do wonders for your work. For the most part, I use an aspect ratio most commonly used in cinematography, which is 2.39:1. There’s no particular reason why I use this ratio, it’s mostly a personal taste, even though it does help to sell that cinematic style since most of us are used to identifying that aspect ratio to movies.
The composition has a much bigger purpose than just making a cleaner and readable image overall, it also helps to emphasize the overall feeling and mood of a scene. Different compositions convey different emotions to the viewer, and it's never a direct or easy answer. There’s a lot of trial and error, some are easier to achieve than others. The main thing is to always keep moving, changing, and never be satisfied with what you have and only call it done whenever it is really done.
Here are some images with several considerations I made when making some of my compositions:
If you want to learn more about composition I would highly recommend a GDC talk by Miriam Bellard from Rockstar North where she talks about Environment Design as Spatial Cinematography. It’s an extremely interesting and packed talk where she talks about composition, lighting, semiotics, movement, and much more.
Approach to Lighting
For my projects, I used Unreal Engine 4, and I only really activated SSGI and Mesh Distance Fields, making it pretty "vanilla". I would also say that I have a much more artistic approach to lighting rather than a technical one.
Since I do not like to use baked lighting because of its long baking times, I always go with dynamic lighting for all of these scenes and I always start with the directional light and skylight and really try to nail them down. I also put a post-process volume early on in order to set the minimum and maximum brightness exposure values to 1 and activate Infinite Extended (Unbound). That way what you see is what you get and there are no annoying variations in exposure if you are flying around with your camera.
I tend to use either custom Skyboxes, Velarion, or the BP_Sky_Sphere that is already in Unreal Engine. Once I’m happy with it, I start putting an Exponential Height Fog and adding other small light sources, I use a mix of spotlights and point lights for different effects. Spotlights in general are nice to create rim light effects while point lights work better as a fill light, but there’s no rule of thumb.
After that, I usually use fog sheets and light beams that you can pick up from the free Blueprints map from the Unreal Engine marketplace, as they are just a quick and cheap way to add fog and light in very strategic and controlled places.
Since I’m only showing a specific composition shot, the viewer isn’t going to view anything else but that shot, so I can use objects outside the camera view in order to get some interesting shadows. For this particular scene, I used a lot of trees for the desired effect:
With that in mind, you also don’t need to set dress the entire scene if the viewer isn’t going to see it.
I also use a cloud function for the directional light on some of these projects, it’s a simple and effective way of creating some interesting shadow and light contrast, you can actually get a cloud function in maps like Rural Australia in the Unreal Engine marketplace for free.
The more you study lighting and check how people do it, the more you realize that most of the time it makes absolutely no sense. Lighting artists end up adding a lot of lights everywhere in order to get more indirect lighting and create more visual interest. On my Medieval House scene, for example, I ended up taking advantage of the small light sources of the windows in order to get much more bounce light.
This doesn’t mean you need to put a million lightbulbs everywhere. It just depends on what you are trying to achieve. In my Barn scene, I kept things extremely simple and the post-process work there is kept to a minimum or even none.
Establishing a foreground, middle ground, and background is crucial for good lighting and composition work. If you check my Pilgrimage scene, it goes from a darker foreground all the way to a lighter background, leading your eye to the goal of the character, which is the door, which also makes it a much more clean and readable image.
Learning color theory is also extremely important for lighting artists. One thing that I always keep in mind is the 60-30-10 rule. It’s basically a guideline that says that 60% of your scene should be a dominant color, 30% a secondary color, and the last 10% an accent.
Textures also play a massive role in lighting and color. Tweaking the overall parameters like specularity, roughness, base color, and others. That’s also why Megascans assets are super cool to play with since they have a ton of parameters you can adjust to match your scene and give artists a lot of room for exploration and make our work more unique.
With that in mind, when set dressing these different lighting setups we have to keep in mind what mood we are trying to convey. The first image has a more serene and peaceful mood to it and the lush forest helps emphasize that idea. The second one has a lot more lack of foliage and its overall colors tend to be more brown and red, which brings a more eerie look to it. The third image brings this idea of desolation and death with a lot more dead forest.
What I end up doing in order to get multiple lighting scenarios faster is that I basically finish (or at least try to get close to finish) one lighting setup, and from there I copy and paste that lighting setup and adjust accordingly. A lot of the lights that you already have can be used again, saving you precious time, and what I end up doing is adjusting the color, the intensity, or even the overall position of the lights. If I feel like they aren’t doing anything to help that specific scene I just delete them or add new lights entirely. The same can be applied with fog sheets and light beams. The one thing I don't transfer is the post-process effect, which I always reset and do from the start for each lighting setup.
It’s also a good thing to keep jumping from one lighting setup to another to avoid looking too much at one of the setups and falling in love with it becoming blind by the mistakes you are making.
I only go into post-process once I have everything more or less settled and I do try to avoid tweaking it too much. I like to put some of those effects that we all know and love like vignette, grain, dirt, bloom, chromatic aberration, and more, in order to avoid the "uncanny valley" effect by adding some of these visual defects. There’s a great talk by Dori Arazi about this in his GDC talk for God of War and how the art team actually used these visual defects in order to make the game look worse.
Unreal Engine 5
The future with Unreal Engine 5 looks promising and super exciting. I have been slowly trying it out and the fact that you don’t really need to bake any sort of lighting to get some great results is amazing. Right now there is a lot of restrictions on how and when to use it, but I’m certain that in the near future a lot of it will change. I’m especially curious about the progress made with emissive lighting. I view these new features with an extremely optimistic look. The fact that we need to worry less about the technical part of the software and pay more attention to the artistic side and focus on delivering interesting and immersive experiences to the player is more than pleasant.
Main Challenges and Advice
Probably the most time-consuming part of doing these scenes is planning: gathering references, understanding what I want to convey in terms of mood and meaning, and also what are the best assets to use in certain scenes. Planning is fundamental and it alleviates a lot of the work thereafter.
I personally don’t like to spend too much time on one piece, since I want to create them as efficient and high-quality as I can, so I establish a deadline of 1 to 2 weeks time with each project. It’s something that you see a lot in the Graphic Design community, where a lot of artists do challenges like "A poster a day" and there is a good reason for it. It’s all about practice, keeping it short and making a lot of errors and experimentation that will eventually lead you to become a better artist and also not to fall in love with the pieces that you do, which I believe is extremely common.
I also picked up a lot of what my mentor Peter Tran does in terms of workflow, which made my own process in creating these scenes a lot faster and more efficient, and I hope I can improve even more.
If I had to give any advice to the aspiring artists in order to create these scenes it would be to practice, practice, practice. Get comfortable with the software you are using, because it doesn't really matter what software it is, learn and constantly relearn the art fundamentals, learn about design methodologies and be confident of yourself and the work that you do. We all suffer from the infamous imposter syndrome from time to time, and that's fine, I think it's good we feel that way sometimes, it keeps us on our toes, as it's a reminder that we need to keep pushing and be better than we were yesterday.
With that I would like to thank 80 Level for the opportunity of sharing some of the thoughts that go behind my work, my IPCA university and teachers like Nuno Martins and Marta Madureira for all the support and guidance, Peter Tran and Ashley Thundercliffe for all the help and mentorship and the awesome discord communities Experience Points, Dinusty Empire and Lighting Bot that I highly recommend everyone to check out.