Rólegur - Stylised Viking environment in ue4

In this article we’re joined by Glen as he breaks down his beautiful scene, Rólegur, which focusses on a Viking being sent to the afterlife under the aurora borealis. Join us as we find out how the scene was crafted, tons of ZBrush tips, stylised texturing in Substance Painter, creating effects in UE4 and much more! We also quiz Glen on life as a Principal Environment Artist, giving feedback, working in a team remotely and that’s just the start!

Intro

Hello all! My name's Glen Fox and I'm currently a Principal Environment Artist at Rare, working on Sea of Thieves.

I studied Games Production at the University of Lincoln, dabbling in most aspects of games development, before landing on 3D art as a discipline. After graduating I spent the following year building a portfolio, while working part time for a medical supplies company, photoshopping a variety of plasters and bandages for their webstore (fun!).

I landed my first gig at Fireproof Studios back when they were an art outsourcer, with my first credits being Little Big Planet 2, Killzone: Mercenary, and Ridge Racer Unbounded-ed-ed. The company moved into mobile games development, so naturally I started focusing more on mobile art, and the more unique workflows and practices that mobile development entails.

From there I spent a few heady years at Space Ape Games; a super-awesome, super-talented mobile studio based in London, before making my way back to the Midlands where I landed a role at Rare.

Composition

I'm still a bit of a newbie when it comes to decent composition, and would like to learn and apply more professional/advanced techniques over the next few years.

I find there's very little videogame-centric resources when it comes to composition and lighting, but a lot of the practices and methodology from other different media (cinema/photography/architecture, etc.) can be applied here.

One of the joys of developing a scene like this is that you have complete control over what the viewer sees, and what angle they see it from. Composition for a production environment is a little more difficult as the character is most likely free-roaming so can peer behind the curtain, so to speak.

For the following shot, I used the common compositional rule of thirds (give-or-take). There is an area of pronounced detail on the left (the hut and brazier, with various brick-a-brack), an area of detail on the right (the longship), and an area of relative calm in middle 3rd.

I try to use action lines to help guide the player's eyes. These shouldn't be too long or obvious, but should be subtle tangents which point towards the focal point of the scene. I actually used very few in this scene, and some are as subtle as the camber of the terrain sloping toward the ship, or the movement of the Aurora Borealis.

Another important area I focus on when it comes to my final composition is usually the silhouette of the focal point itself. The area around the ship is less detailed, and helps create a clear silhouette around it. There are no adjoining tangents, or things cutting through the focal point. This clear silhouette is coupled with a value separation, to help make the focal point 'pop'. In this case, I decided to frame the ship with the curvature of the distance cliffs, to create a slightly more forced/obvious framing.

I have a tendency to want to add on ALL the details, but I tried to restrain myself a bit here, adding most smaller props around the left 1/3rd of the frame, rather than crowding the area around the ship. Clusters of smaller props can really help sell scale in more stylised scenes, and can be used for environmental storytelling, so they're important.

Colour and Light

As with composition, lighting is an area I would like to level-up over the next couple of years. Once again, a lot of the techniques from other media such as movies and photography can be applied here.

There are, as with composition, a few key considerations which inform the lighting direction of my final scenes. The most important consideration is the use of values to guide the viewer's eye.

I regularly take screenshots while working on a scene and take these into Photoshop. Setting my 'working proof' to grayscale helps me to quickly see light and dark values in my scene, and check that there is separation between the important elements. I can then course correct accordingly.

The area around the ship is much lighter than the ship itself, in order to make the ship pop. Aswell as this, the area around the edge of the frame is darker than towards the centre of the frame, to gently direct viewers attention inward.

I also regularly look at the shot in thumbnail form, both in greyscale and lit, to make sure the focal point remains, and no unintended areas are grabbing attention. I quite often find my face planted way too close to the monitor to take the scene in as a whole, so occasionally stepping back and viewing the scene zoomed out like this really helps.

I spend time early on making sure that my materials are crafted to work well with the scene. For this scene I used muted colours and cooler hues, so the blue-tinted lighting would layer over them well. I wanted the sails to be more striking, but I knew they were going to be bathed in a warm orange glow due to the fire, so I made them a light, neutral colour.

Because materials have such a radical effect on the values in the scene, I generally tend to block these out early on and keep tweaking the albedo hue/saturation/lightness, until I find an overall colour palette that works and helps blend everything together. I rarely fully texture an asset or drill down into the minutia of surface treatment until the majority of the scene has a flat colour applied and I'm happy with how everything's sitting together.

A lot of the overall colour in this scene comes through post processing, in this case a LUT. Much as in cinema, overall grading can tie everything together nicely, and can also greatly change the perceived tone of a scene.

Creating a LUT in Unreal is super easy. I take a screenshot of my scene into Photoshop, and start going ham with some image correction layers until I find the look I want. These correction layers can then be dragged over the default LUT, and exported to the engine. I find colour correcting much easier in Photoshop, and the beauty with this process is that what you see in Photoshop, is what you'll get in-engine.

For the post processing here, I added some blue in the shadows, which makes the scene feel cooler. I also slightly increased the contrast, and played around with the vibrancy of the scene (it's much easier toilet-nudging things like this at a global level, as opposed to on an asset-to-asset basis).

I try and do this, along with other post processing tweaks, when I'm getting pretty close to completion with a scene. I personally find that if there are more variables being tweaked at once, then things can get convoluted very quickly. You often can't remedy any colour or lighting issues as easily: 'Is that part of the mesh too dark because of the material, the lighting, or the grading?'

Lighting and Camera Setup

For this scene, I wanted to create the feeling of a cool (temperature) moonlight, transitioning to a warm glow from the fires.

The main light is the cool blue directional light of the moon. There's nothing very fancy about this light, but it's the most important one in the scene. I spend quite a while towards the start of whiteboxing, constantly tweaking the light's rotation and brightness, to make sure I'm happy with what parts of the scene are in shadow, what parts are bathed in light, and with how much intensity. I find it too easy to oversaturate a cool light, which can make it feel artificial, or overblow the light, which also makes it feel artificial.

There is another directional light emulating a minty colour cast from the Aurora Borealis. I thought this would be a good way to get some rim lighting on areas of the hut and totems, to help the silhouettes to pop a little more. Disclaimer; this ended up a lot subtler than I had originally planned.

I wanted the light for the fires to feel a lot warmer than the surrounding area, which also helps give contrast between these areas of interest and the surrounding areas, so you have the change in value, and also the change in temperature.

These lights are simple point lights with their attenuations adjusted and larger, softer sources. They also have a simple light function applied to them, to emulate the flicker of fire.

A lot of the bounce lighting comes from a skylight with it's indirect influence boosted. There are also small pointlights with soft sources and very little specular influence, placed in areas where I would like a little more (fake) bounce light.

The camera for the scene is a simple cine camera, which uses a sequence to move between two points for a slow pan/zoom. I'm a bit of a layman when it comes to camera settings, so I generally select one of the premade lenses for the camera, and then just tweak the aperture settings to suit the viewpoint and scale of the scene.

The camera has a slight depth of field effect after tweaking this aperture, which helps to slightly soften the finer details in the foreground, and push focus to the midground. It's pretty subtle, and I find that if I push this blur too much, it can greatly affect the perceived scale of everything.

Particles and VFX

I knew from the outset that I wanted to record a video for this scene, in order to capture the Aurora in motion. so wanted to make the scene as a whole as dynamic as possible. Any extra ounce of movement I could squeeze from the scene the better.

The trees have two levels of movement to them. At a more micro level, the branches wobble (pretty unconvincingly) by using a vertex colour passed through a simple grass wind node in Unreal, connected to the world position offset. It's pretty unrealistic, but enough to give the notion of the branches being affected by wind.

At a macro level, the tree bends using a material function and a global wind direction parameter. This is masked by the world position of the tree (the higher it is, the more it will bend). This wind bend function is covered in CodeLikeMe’s videos, which was where I learned the original method. I wanted these trees to feel pretty tall and top-heavy, so having them sway gently in the wind was important to me.

The water covers a decent chunk of the shot, so I spent a while focusing on that, and experimented with a few different styles to see what fitted the scene better in terms of stylisation. I ended up with a more cartoony wake and wash, but with nice soft flowing reflections.

The water takes a lot of cues from games such as Rhyme and Zelda: Wind Waker. In fact, Simon Trumpler of TequilaWorks has a fantastic talk on YouTube covering how they approached a lot of the VFX in Rhyme. Well worth the watch!

The water uses depth fields to detect where objects intersect it. A panning noise map is clamped and applied to these areas, in order to get the stylised wash. A soft bumpy normal map is then panned across the surface at two different speeds (then blended together), so you get some nicer, more organic reflections. I wanted the water to be quite stylised, but still wanted the reflective properties and finer ripples of a more realistic water surface.

It’s worth noting that I didn’t bother doing any fancy translucency or refraction to the water, as I didn’t think I needed it for this particular scene, and it made the shader much less complicated to make, and also much cheaper!

The wake uses the same exact method as the wash, but is mapped to a mesh so it’s easier to direct the flow of the water, and works without meshes having to break the water’s surface.

The snow particles are very simple emitters that spit out small round, white particles. These particles are affected by gravity and with a vortex deformer in Niagara, in order to give them a more flowy effect, as if they are being whipped around by the winter's wind.

The fire emitter is again a pretty simple Niagara emitter, with colour and alpha changes over life, along with some rotation and velocity deformers. There are also some small embers to give a bit more layering to the fire, using the same simple deformations, but with different velocities and lifetimes.

A few more, larger particles are emitted underneath, with a simple noise normal map and their refraction set fairly high. This creates a nice heat distortion effect. This is very expensive, but gives a nicer sense of heat coming from the fire, so would probably be used sparingly in a production environment. Disclaimer: this effect is more pronounced in-motion than in screenshots.

Sculpting Techniques

After spending so long sculpting assets for Sea of Thieves, I think a lot of the stylisation and general sculpting principles have translated to my personal work (still not sure if this is a good thing or a bad thing).

I generally keep my sculpts pretty simple, with only a small amount of noise or surface detail, and focus on creating nice, interesting, and readable silhouettes, while describing the surfaces/materials of the sculpt just enough.

When it comes to the actual process of sculpting, I use the following 3 brushes for the majority of my sculpts, along with the good old faithful clipping tool.

I use the clipping tool to create crisp bevels and cuts. This can be really destructive to the topology of the high poly mesh, but creates some nice crisp and defined edges in a couple of clicks. These can be softened with claypolish if they're a bit too harsh. Ctrl+Shift starts the cut, with a gradient showing which direction will be sliced, and a double-tap on Alt puts a hard angle in the cut, which is crucial for the sharp edges.

I also use booleans and dynamesh quite a lot, for more complex or overlapping shapes. I try to break more complex shapes down into simpler shapes/parts and then work these separately, before bringing them all together and dynameshing. For instance, the curly details in the examples below were made from a few separate subtools, then combined. Working this way can also be crucial when wanting to sculpt concave shapes while keeping smooth, clean facets with crisp edges.

For most assets I like to start by creating a proxy mesh in Maya, which I can then export to Unreal and dress the scene with. This allows me to see the asset in context very early on, making sure I'm happy with overall proportion and feel, and then I can replace the proxy mesh directly with the final asset at a later date. This proxy mesh is also then taken into ZBrush and is the base that I start sculpting on. I know some artists prefer to start from scratch in ZBrush itself, but I personally like to go as far as possible in Maya with using poly modelling, subdivisions, and the crease tool, as this is the package I'm most comfortable with.

The amount of stylisation applied to the meshes depends on the look I want to achieve for the scene; more bowing and wobbly shapes can often lend to a more cartoony look, while straighter edges and thinner elements can create a more mature look. For this scene, the surfaces have a bit of wonkiness and interest to their silhouettes, and slightly chunkier elements to them, without adding any tapering or bowing, resulting in a stylised look that’s not too young or cartoony.

I like to think about this shape language while sculpting my first few props for the scene, making some basic rules around how to approach different surfaces, before applying them to other assets early on. This helps create consistency. I was pretty lucky with this scene because I only really had to think about my approach to a few big materials, since Vikings were pretty limited with what they had available to build with!

Texturing

I try to blockout my materials before fully texturing them, to get a good idea of what sits well in situ in the environment. This means that the first iteration/pass for the scene is to apply a flat roughness/metallic value, and a flat albedo value to the larger assets/surfaces in the scene (the ones which will have a large effect on the overall feel of the scene). I then start to work more on the larger surfaces or hero assets, but only after the base pass is pretty much completed.

The small minutia of a texture can add a lot of complexity and depth to individual assets, but it’s amazing how close you can come to a fully realised scene with simply flat albedos and metallic/roughness values.

It’s worth noting that I use the buffer visualisation channels In Unreal regularly, as it helps me balance the separate aspects of the scene materials in isolation, e.g. albedo, roughness, etc. Obviously the final, lit materials are what’s most important, but checking an individual channel can often help when remedying imbalances across materials in the scene.

When I'm fairly happy with the blockout texture pass, I'll pick a couple of items which incorporate the larger or more abundant materials in the scene, in this case the hut, longship, and cliffs. I'll then take these and start working them up a bit by baking the high poly information to them inside Marmoset Toolbag, and then taking them into Substance Painter to texture.

Marmoset Toolbag has become my package of choice for baking out all of my high poly information to the low poly mesh, due to how quick, easy, and iterative it is to use. I especially love the fact that you no longer have to explode meshes to bake them. It’s also a very fast baker that produces clean results.

My workflow in Marmoset Toolbag is pretty standard so I won’t go into detail about it, but the maps I usually bake out are; normal, curvature, AO, world-space normals, and (if needed) alpha and height maps. I then take these bakes and the low poly mesh into Substance Painter.

The first thing I do when opening Substance Painter is to set up a custom HDR environment. I used to always texture assets using one of the inbuilt ‘studio’ environments, but found that exporting the assets to the engine then left me with radically different results, because the lighting itself was radically different. I usually either just export an environment map from my whiteboxed (with flat colour values and blockout lighting) scene, or alternatively find a HDRi online in the ballpark of my ideal tone and lighting for said scene. Accurate reflections don’t matter to me here, it’s more important that I have an environment that resembles the lighting colour and contrast of how I want my scene to look.

I generally don't do a lot of granular details in my textures, focusing mostly on a few key attributes - edging, crevice dirt, baked AO, and colour variation. I use my baked AO and curvature textures as masks for a lot of these, with some blurring and levels to make sure the masks are nice and chunky. Clamped perlin noises and some generators (with any ‘grunge’ stripped out) are used to break up these masks when they’re too clean or precise. Colour variation is usually achieved with a group of layers set to overlay sat on the top of the material stack. These layers then have clamped perlin noises for masks.

Once I have a benchmark asset, with materials that I’m happy with, I’ll convert them into Substance Painter master materials and carry them across to the other assets (tweaking any masks or values to get the best results on each asset). This saves a huge amount of time, and helps with consistency.

Aurora Borealis Effect

I tried a few different things for the Aurora Borealis, before settling on the following technique. The mesh itself is an s-shaped plane with a fair few vertical cuts. I did quite a few cuts to make the movement smoother (I also knew I would only have a few of these on-screen at a time so polycount is less of an issue). A panning displacement map is used to create the soft undulations displacing along the surface. 

The vertices on the end of the mesh are painted black, which is multiplied to the alpha channel of the diffuse, to create the final opacity, and the fade out.

The diffuse itself is a pretty simple array of softened dots of colour, with some airbrushed gradients sitting behind these. This diffuse is horizontally panned and when combined with the mesh undulation, creates the organic swirling effect.

Everything is set to be additive so the effect never appears duller than the sky behind it.

If You Had To Start You Career Again?

I honestly don't know if I would change much really. I've had the privilege to work with some amazing people on amazing projects, creating some content that I'm very proud of. I've made some lifelong friends, had many exciting adventures, and managed to avoid any of the more toxic or stressing things that unfortunately exists in small pockets around the industry.

I would definitely tell my younger self to be less reticent to finish projects, and to always take pride in what I produce. Anyone who knows me knows that I like to start 10 things and finish 1! A lot of this probably stems from a subconscious feeling of not always being 100% happy with the direction of some of my personal projects, and not being motivated to finish them, because I don't think they're good enough. It's a pretty terrible trait to have, but one I know a lot of artists suffer with.

What Separates Junior, Mid and Senior Environment Artists?

I think you would get different opinions on this, depending on the area of the industry you're talking about, at which studio, and on what project. However, here is my personal view, speaking from my experience, in the roles that I've held or gravitated towards.

There are a few key things that seperate junior/mid/senior roles. The most obvious, and overarching differentiation between a junior and mid-level artist, (and the one everyone seems to acknowledge the most), is work experience and technical expertise. I'd agree that this is probably the main separator between a junior and mid-level artist, but there are a few key clarifications beyond that when you pass mid-level.

Senior artists are expected to be a bit more knowledgeable about things such as actual development workflows beyond art, technical limitations etc., and not just more knowledgeable with the software and general content creation. This comes naturally as you work in development more, and (generally speaking) education won't prepare you properly for this. On-the-job experience is key.

Senior artists are also expected to look at the bigger picture, with a greater emphasis on proactive problem solving, and going the extra mile to help improve processes and streamline workflows etc. A good example may be to work with tech art to standardise or streamline the exporting of certain assets, or perhaps working with other artists to help create a library of typical ZBrush assets for quick and easy kitbashing on a project.

My main piece of advice to anyone looking to progress in their career is to apply themselves beyond just completing the tasks that are handed to them. Ask questions, follow things through to completion, and definitely own tasks wherever you can. When I say 'own a task', I'm talking about seeing a feature through from when it gets handed to you, through to when it's in the game and in the player's hands.

Working on your own art projects outside of work will naturally improve your artistry and software knowledge, and for a lot of artists is a great way of invigorating their creativity. However, when it comes to practical application in a development environment, it’s important to have the more development-centric skills aswell. I like to compare it to the masterpiece of cinema that is Twins. Danny DeVito and Arnold Schwarzenegger are both fantastic in their own rights, but it’s only when the two are combined that it really elevates it beyond each counterpart (most of my metaphors and life lessons relate to Danny Devito in one way or another).

Principal Environment Artists

This will no doubt differ depending on the company you're at, and the project you're on. For us the biggest aspect is to help newer team members adapt to the Sea of Thieves way of content creation; our pipelines and workflows, technical limitations, and (probably most importantly) the art style.

We spend some time helping them adapt to our art style by actually creating some content for the game, while providing periodic feedback and paintovers, and talking them through any processes or ways of working that they may not be used to. Working remotely makes the whole process slightly different as we can’t just check on someone sitting at a desk, but the general onboarding flow remains the same as if we were in the office.

Because the project is so established now, we have a tonne of great documentation, and many well-versed artists on the team, so onboarding new artists is a breeze, and something that is shared across the team.

Best Way To Give Feedback

Different people prefer to give feedback in different ways, and different artists are definitely more or less responsive to different forms of feedback. This is important to bear in mind, as it'll change how you should approach feedback, or whether or not your feedback gets across effectively.

Personally, I love giving feedback via simple paintovers as it helps demonstrate points much easier, and more importantly, it also proves out the feedback with just a small amount of work. This means the artist will actually see what effect the feedback will have on the asset/scene before deciding whether to implement, and actually doing the implementation itself. This is a more visual way of feeding back, and should also help the artist understand your process better.

Before the obvious COVID restrictions, there would be a lot of face to face discussion around people's desks, and you would often find artists taking a break and checking out what other artists are working on, striking up ad-hoc discussions. With the current restrictions these opportunities are gone, but we can still do things like call each other, do a cheeky screen-share, etc.

What Do You Think Could Be Done To Help New Team Members or Keep a Team’s Continuity/Comradery Strong, Especially When Working Remotely?

One thing that I think is really helping out our team is making sure that we encourage a mindset of growing, learning, and sharing, while giving the artists a forum to do this. It's usually as simple as having a channel on Slack/Teams/whatever you use to communicate that’s a general ‘art’ channel. This generates a lot of chatter between the artists and helps keep them motivated and stimulated beyond the tasks that they're working on. It also works as a great way to give and receive feedback, and get people's opinion on things.

Regular team catch ups are  also crucial. It's all too easy to work siloed at the moment, meaning artists stop collaborating, stop asking each other for help, and just generally stop communicating. Our Lead Artist is fantastic at making sure we don't slip into these habits.

How Do You Ensure That Your Team Maintains The Pipelines and Art Direction When Working On A Project?

As mentioned, we have a lot of awesome documentation which heavily outlines our art direction, along with some artists that are well versed in creating Sea of Thieves artwork. This art direction and style ruleset is instilled in the artists’ minds from the moment they step onto the project, and kept regularly in-check by way of things such as art reviews, and general peer-to-peer feedback. The artists are encouraged to take pride in what they do, and to think about benchmarks for their art.

The quality bar for the project always has to sit at the back of the artists' minds. Timeframes can often be tight, but that doesn’t mean that quality of the artwork or the implementation should suffer. Having to go back and fix things that could have been done better the first time around can often end up taking longer.

I think that by giving autonomy and ownership to individual artists, it definitely helps maintain a high and consistent quality bar (this empowerment is something that Rare is great at). If an artist owns a feature, they're going to be more inclined to ensure that they see the feature through until it's in the player's hands, and they're more inclined to make sure that the feature is at quality. It also helps them to be more proud of their work 'Hey, check out this feature that I helped bring to the game!'.

If A Team Member Is Struggling Working On A Prop, What Do You Or The Team Do To Help?

I feel like I'm extremely lucky in this regard, as I'm part of a team that smashes it out the park consistently!

Generally, if an artist is struggling on an asset, they may reach out to the team for help. We have a very fostering environment in our team, so I don't think anyone is afraid of asking questions, or seeking feedback.

It depends on what type of feedback that particular artist is more receptive to but the team will usually help out with bullet-point feedback, paintovers, practical examples over screen-share, or maybe just a meme or two. If the feedback is unclear, or the problems are easier remedied ‘face-to-face’, then we’d definitely call or screen-share and talk through the steps. It’s hard to not be able to stand at someone’s desk and point at a screen, but we make do in these crazy times.

What Does Your Average Day Look Like and How Does This Differ From a Lead Artist or Art Director?

I think this will, once again, differ depending on the company/team you're on, and the project you're working on. I'm personally part of a smaller art team, on an evolving, live game. Bearing that in mind, the following is a snapshot of what a typical day may look like for me.

Being a Principal Artist at Rare is still a very art-centric hands-on role, something I'm mega grateful for. This means that a large portion of my time will still be used actually creating artwork, and not in various spreadsheets/jira boards.

An outsource drop usually comes in during the morning, which involves checking through the artwork and maybe providing feedback, paintovers, or integrating the assets etc. We've gotten to a great place with this and have the privilege of working with some very talented external partners, which means that I usually spend this time 'oohing' and 'aahing' at some great looking artwork.

A few meetings may be scattered throughout the day, whether it be art reviews, discussing various game features, company-wide updates, etc. Artists at Rare have a great deal of self empowerment which means that different artists own different features. With this more flat hierarchy, it means less meetings (as the feature owners only go to their relevant meetings), and more time to make art!

The lion-share of the day will be used to create any in-house art for upcoming features, supporting the other departments with any smaller tasks they may need, and dealing with any bugs that may come in from QA.

Beyond this, there are a few nuggets of time spent on things such as task management, and drinking my bodyweight in coffee.

The difference between a Principal and Lead role is quite a significant step up in terms of responsibility and non-art workload, while I view the role of an Art Director as a more lateral step as much as a step up.

The Lead Artist spends a lot more time rallying the troops, making sure that art comes in on time and that the team is aligned on objectives. They will also have a lot more higher level conversations, and are generally a conduit to the management/other departments. If there are any big conversations about art requirements for upcoming releases, they’ll be there.

The Art Director is essentially the vision-holder for the project. They have the responsibility of making sure all art departments are in line with this vision (and therefore the art style of the game), and are expected to have eyes on the game as a whole package, rather than just the separate elements. They also (like the Lead Artist) spend a lot of time in super-fun management meetings, figuring out the logistics of bringing certain features or updates to the game, and making sure that departments are talking to each other.

Where Are You Still Looking to Grow as an Artist?

I think one of the most important qualities in becoming a better artist is self awareness, which means being able to realise weak points and working on them. I realise that lighting and composition are probably my weaker points when it comes to the grand scheme of things, and the things I'm most interested in improving, so I'm trying to create some bigger scenes which will hopefully let me flex these muscles and help them grow.

At a more software level, I want to work more with Substance Designer, since it's such a powerful tool that's used heavily in the industry now, but still remains a bit of a mystery to me. I know how to use it, but a lot of my time spent in the package involves me fumbling about and re-remembering what buttons do what, and how to achieve certain results.

Inspiration

I think there are probably two main sources where I get most of my inspiration from. I find scrolling through Artstation or Twitter is a great way to find artists/artwork to aspire to, concepts to work from, and art styles and techniques to try. I do however personally find the line between inspiration and self-destruction thin when scrolling sites like these, so I'm trying to limit their usage a bit. I find a lot of inspiration here, but it's all too easy to start feeling insecure, inadequate, and overwhelmed by all the amazing artwork floating about nowadays.

I also naturally get lots of inspiration just from consuming vast amounts of different media.  Playing games, watching movies, and reading artbooks can rejuvenate me and make me want to start even more projects that I won't finish! I also find that playing games can help with projects a lot, because you get to see how other artists have approached certain things that may be related to something you're doing yourself.

I'm not ashamed, but I'm one of the people who will play a game just because it has pretty rocks in it that I can stare at. #rockchat

Feedback

I honestly find myself seeking less external feedback nowadays. This isn't to say I don't need or want feedback, far from it. I just seek more focused feedback from smaller targets, away from some of the larger forums or art groups. If I'm looking for feedback now, I'll reach out to artists that I admire or who I would naturally gravitate toward artistically.

This is a great way of not only getting more laser-focused and personal feedback on what you're working on, but making sure that the feedback will guide yourself and your artwork toward the art style and benchmark that you want to achieve. Don't be afraid to reach out to artists; some are busy and won't reply (don't take it personally), some reply months later (guilty), and some will reply with amazing feedback that will blow your art-game wide open.

Tips and Advice

Be critical about your own artwork but appreciate your own growth aswell. Self reflection is one of the greatest traits to have as an artist. Always ask yourself, ‘If I were to make this piece of artwork again, how could I improve it?’. What were the things you think went well?  What didn't go well? What did you learn from the process?

If you're aiming for a specific position in the industry, at a specific studio, then start tailoring your artwork towards that. That's not to say that you should pigeon-hole yourself, but crafting pieces that closely fit with the IP or company that you want to work for, will show them competence creating content that would work immediately with their game, and also shows your passion for their particular brand or ethos.

Finally, I'd definitely say that (with the previous point still in mind) you have to create artwork that you're passionate about. If you're not enjoying creating the art for your portfolio and aren't proud of what you're doing, then you won't enjoy doing the same when you're on-the-job (especially because it’ll come with a side of bug-fixing and optimisation).

Future Works

I have a couple of scenes that I'm tinkering with at the moment, experimenting with different themes and slightly different art styles. I'm trying to go outside my comfort zone by incorporating some different software and ways of working, including larger scenes, with more complex lighting and composition.

I also have about a million other projects I’ve started, which I may complete at some point between now and my 100th birthday.

Outro

Thank you for reading through all of my ramblings! I hope you've found some of this enlightening, and have a better view now of how I approach my scenes. 

If you want to see more of my work, or have any questions, feel free to reach out and follow me on Artstation, Twitter, or Instagram.