Concept Art Switch from 2D to 3D

Ned Rogers talked about his first experiences with Blender and explained how 3D modeling tools help him create better concept art.

Intro

My name is Ned Rogers and I’m a concept designer from Adelaide in South Australia. I’ve been working in the entertainment industry for the last 6 years or so on a variety of project across kids TV, video games and animation, mostly as a freelancer. I’ve been lucky enough to be able to work for a bunch of different people and companies: Wizards of the Coast, Technicolor, Flying Bark, Bosskey – it’s a fairly varied list! Currently, I work as an instructor at CDW Studios in Australia which is a school focused entirely on entertainment design where I teach environment design for games and vehicle and mech design.

Switching to 3D

I’ve dipped my toe into using 3D for concept work on and off over the years – mostly just to figure out some tricky form or perspective problem but I never really found any toolset or workflow that held my interest for more than a couple of jobs. Everything just took too long or required too much forward planning that I couldn’t really design and model at the same time. There are some people who do a good job at that and have been doing so for years but I was never one of them!

So I went through a few programs – ZBrush, Modo, Sketchup – but just ended up putting them back in the cupboard after I had struggled with them for a while.

I had known about Blender for a while, it had always been in the background being championed by a bunch of people who I assumed were fairly sadistic as it just seemed like such a weird program to use. But then, I was lucky enough to meet Vaughn Ling at a masterclass he ran at CDW: he was showing us a bunch of stuff he’d been developing in Blender and a set of plugins that were geared specifically at quick modeling for the concept. The Heavypoly toolset seems to be pretty well known now, but back then I hadn’t seen anything like it – and especially not attached to completely free software. So that was pretty much it for me, I did a couple of tutorials and in a few weeks, I was able to work in Blender at a decent level. Then, when the real-time Eevee engine came out, not even having to wait for renders made it even more attractive.

Being able to build modular assets, arrange them as a set and then explore multiple angles is invaluable and allows to give a client some options to explore. Once you have a good setup and a few reusable assets in your library you can start to kitbash your own designs and work to produce variations really quickly.

Advantages of 3D Concepts

Flexibility is probably the biggest advantage of 3D. Drawing turnarounds isn’t a thing since you can just render a few angles and any late changes won’t require a full re-draw, just re-render the angle and you’re done. And from an environment and set perspective, being able to tweak your cameras and find new shots is really cool. Of course, there is nothing new to people who have always been comfortable with 3D, but for me, having it attached to such a fast set of tools is really cool. I’ve also found that being able to set up cameras at a gameplay angle is a really valuable tool for making sure you’ve designed an interesting space for a person to be in, not just something that looks good as a concept painting.

Working with Blender: First Experience

I did find Blender a bit weird to start with but mostly because it just has a lot of tools and functions. Once I started using the Heavypoly scripts, it became much easier since you can only look at the tools you need to quickly create volumes and move them around. The rest is moved to the back so it doesn’t distract you!

Having had some experience instructing students who are often new to professional tools I find that this is actually the thing that trips beginners up as well – they just get overwhelmed with the number of options and tend to jump in and start throwing stuff around. It can be a good way to learn which buttons to press but it usually results in pretty things that have a lot of bad design behind them.

I think it’s really important to have a clear visual goal when learning any new tool, so if you do a study of something that has already been solved from a design point of view you can concentrate on making it well and learning the new tool. Then, once you’re comfortable you can move on to trying to create your own designs.

Efficient Modeling in Blender

I don’t know if I’m the right person to talk about efficient workflows, because I keeping up the speed, I make bad models that any 3D artist would probably feel ill when looking at. But from the concept side, that’s also probably the answer to how to be economical. At the end of the day, I know that my 3D assets aren’t really going to be used by anyone else so I can purely focus on whether or not it looks right for me. I think it’s also important to decide early on if you’re going to just very quickly model something simple and then do more painting over the top or if you want to spend a lot more time in 3D and then work on it in Photoshop for an hour. Getting stuck in the middle is usually a waste of time and you end up fighting yourself.

From a technical point of view, I rely a lot on booleans and procedural textures, which means I don’t have to UV anything and can just leave my topology really rough in the scene. Having booleans that are live and editable is really handy, too.

Lighting

Eevee has a really nice lighting built right in so I don’t have to try very hard to get it looking nice. I’ve always been interested in photography and cinematography so I guess I’ve taken a bit of that across to Blender. Using a simple cinematic lighting setup gets good results and with the volumetrics, you can get a nice shiny ‘concept art’ look with just a few clicks. I use an HDRI to get nice world reflections and Eevee’s screen-space reflections do the rest. When I move into Photoshop, the fact that the lighting is already resolved means I can quite easily work from the palette in the image to paint in extra details I need. I take a cavity pass and a clown pass out of Blender along with the combined lighting pass and that’s it.

Conclusion

I think there’s plenty of room for me to get faster and smoother with my modeling and I’ve been currently trying to work out the best way to push the stylization present in my drawing into a 3D space. I’ve mucked around a bit with a few VR 3D programs and there’s a bunch of cool stuff you can do with those. The challenge for me is having the patience to stick with a slightly slower process when my brain wants to jump ahead to the next problem but I think that’s just a matter of time. I’m probably never going to stop needing a pen and paper for really quick design but it’s really exciting to see where the real-time 3D production can lead.

Ned Rogers, Concept Artist

Interview conducted by Kirill Tokarev