Blacksmiths Shop Revisited

Have you ever had a project you’ve looked back on and thought how you could’ve done things better? Well Paul took it one step further and completely redid his Blacksmiths Shop from over 2 years ago! Learn how Paul created his scene, what changes he made and how he managed to improve on his previous version.

Intro

Right, my first proper article, don't mess this up, "Hi" (so far so good). My name is Paul Carstens and I'm a 3D artist, specializing in props and environments, originally from South Africa but currently living in Toronto Canada. I've been knocking about the 3D industry for about 6 years now but got into the gaming industry proper about 3 years ago and it's been a hell of a ride so far. Back in my day (Someone please help, I'm nearly 30 and terrified about it) studying anything related to making games wasn't really an option but luckily there is a digital animation college, in Cape Town where I grew up, offering a 3 year diploma in animation. So if we want to get technical I'm actually an animator (I'm not an animator) but environments have always been my focus and passion. EXP originally approached me quite a while back about writing an article for them but at the time I didn't feel I had anything really worth sharing, I mean I had done some "show-off" projects for my portfolio but nothing I felt really warranted an article. However, I did mention I was re-doing one of my first portfolio projects as a kind of "HD" remake where I would use all the skills I had picked up and grown over the past 2 or so years since I originally posted it. Now this is a project I feel I can write about, a fun subject and it's always cool to critique your own work with 3 years retrospect.

Original To Remake

I'm going to try and be brief and selective with the points I choose to focus on otherwise we'll be here all day.  So here we go, first here are a few side-by-side comparisons of the two projects:

The original idea or concept for the scene came from a reference image I found of a tabletop miniature for a DnD type situation (miniatures are awesome ref for modular stuff by the way, seriously go check that out), the image was of a little cottage type structure much like I have made. But more importantly it had an anvil and forge and that meant I could make a bunch of medieval weapons and props. That's all I really wanted to achieve at the time, a bunch of cool looking middle ages stuff thrown together in UE4. In all honesty I had very little clue what I was doing in UE4 at the time but more importantly I hadn't quite made the discovery of how vital feedback is to an artist working in a consumer driven industry. If I had not been so "precious" about my work it would have turned out a whole lot better, maybe not great (I was pretty green at the time) but undoubtedly better. I mean just look at all the differing scales in the scene, giant door, huge stairs and a variety of props ranging from tankards for toddlers to axes so big no man could ever lift them and lets not get started on the lighting fiasco.

Now with the benefit of a few years experience and a real conscious effort to seek feedback I wanted to have another go at it. The main focus for me this time was to keep a consistently high quality (both technically and artistically) and with any personal project I do I always want to learn something new. A new software, workflow or even art style, as long as there's something. This time the big one was learning to photo scan with Meshroom (super fun, easy and best of all free), thanks to this I managed to use pretty much all my own foliage and scans not to mention learnt the process from taking a bunch of images using a light-box to retopologizing and cleaning the raw textures. (No more hours spent sculpting bread for me!)

More than anything though I wanted the chance to show myself how far I had come. Impostor syndrome is real, particularly when making games for a living has always felt like a pipe-dream for someone from a country where the industry isn't all that present. Working on personal projects like these is my way to keep the dreaded feelings of "I don't belong here" at bay (for a little while at least) and I highly recommend it. As a side note it's also highly therapeutic to be in full control of a project where you technically don't have to listen to anyone (You really should though, feedback is everything). As artists we are a passionate bunch and after working on a studio project all week where you usually won't be getting to have everything done the way you'd want, it's good to know that you have somewhere else to spend that creative energy. No one wants to work with a difficult person and yes being passionate is pretty much essential but knowing where and how to compromise is just as important when working in a team environment. Projects like these also give you the opportunity to work in an area of game development you usually wouldn't. Thanks to the aforementioned animation training I first got started in 3D as a character modeler and rigger, which I don't get to do so much these days but using those skills for personal work helps keep them more or less up to date. In this example, the set of knight's armour. With the original project I only made a helm (and a rubbish one at that), I think this was the weakest asset of the project so with the update I wanted to change that more than anything. This time around thanks to a ton of reference and feedback points I produced what I would now consider one of the strongest pieces of the project.

Now the armour isn't quite the best example to illustrate my point of why skill diversity is important, beyond a bit of character work, so I'll use an asset from my Feudal Japan challenge on Artstation. The concept called for two "Foo dog" statues, all the references I found had asymmetrical elements but this meant that I would have to pick a pose and be mostly stuck with it if I had just sculpted the statue that way. But with a basic rig I was able to sculpt the Foo dog in a neutral pose using symmetry and then play around with poses later until I found one I liked. After that I would drop it back in ZBrush and apply final touches. In my experience this is a far better way to approach assets like these because it's mostly non-destructive and quicker to produce and iterate upon. The final pose chosen was pretty simple but having the option to change it if I wanted to lets me relax a bit about the whole thing. In a production setting: let's say your lead or AD wants a different pose then there's no problem, you always have that rig stored and can do as many versions as you like. Rigging with z-spheres in ZBrush is also an option.

Composing The Shot

With this project I didn't want to think about composition in the same way I would for a more cinematic piece, being an environment artist for games means that you usually have very little control over where the camera is because the camera is the player (and players do very unexpected things and can come from many angles) So the idea I followed here and do in pretty much all of my "work" work is a piece of advice that I'm sure many of you have heard by now but I first heard from the awesome Martin Teichmann (Now Senior artist with Blizzard) during a Gnomon talk he did back in 2016. He was going over his environment work on Uncharted 4 and made the point that (and I'm paraphrasing here) you want to have "pockets" of interest, groups of props like islands in a sea of resting space. A bunch of assets scattered about evenly just becomes noise, you don't know where to focus and on a technical note it'll probably end up being less efficient in terms of performance. With clusters though you have the opportunity to set the scene a bit and tell a little story which is great for environment art and environment artists, having a story in mind when populating a level means I have a point I'm working towards and know which assets I will need to make or re-use. Much better than just dropping a bunch of generic "stuff" in to fill the space.

For example, the workbench next to the forge. Initially I didn't have a plan for what I was going to have happening here but once I had finished the suit of armour I really liked the way that light interacted with the shapes of the gauntlet, hands are interesting in general so I decided to base a little "composition" around one. The idea here was that it is perhaps the last piece the blacksmith was working on to complete the set, so he's got some etching and painting tools, some spare rivets and a few hammers and mallets near-by. This scene also gave me the opportunity to art direct the lighting a bit more, essentially I have a more cinematic light setup here to bring the areas I want emphasized into focus. Using UE4's light channels I can give some assets a stronger rim, kick or key light without affecting all of the scene around it. (warning: this may not be highly performant, artist's desecration is advised) This light linking may not be physically accurate but (I'm going to drag yet another artist's name into this) Boon Cotter (lighting artist, Naughty Dog) did a talk to introduce his course with CGMA where he mentioned how bending the rules for lighting is necessary to get the look you want and to make sure the player has the experience you want them to have. After all, commercial games are a product and the players are the patrons. Unless you're making a tech demo for super accurate lighting, gameplay should be king. My role here was to know when and by how much to break the rules and get the look and focal points I wanted. Having the gauntlet missing from the rest of the set on display is also a fun way to get the player's attention and lead them to where you want them to go, the armour itself is a focal point thanks to the fact that its the closest thing to human in the scene. So with a missing piece the theoretical player may notice this and wonder, "where is the other hand?" then go looking for it and discover where it is and hopefully the other scenes that are about the place too.

Now we move to the collection of assets at the front door. The boots with mud on the soles, the bundle of sticks near-by, the tankard and bowl. All these can tell the player a story: where did he go? Did he go collect those sticks? Why? Did he stop for lunch and a drink? Where did those leaves on the floor come from? Is that sage hanging up there? Why is there sage hanging up there? Is it for cooking or is it a superstition thing? The hope is that this all helps a virtual world feel alive and lived-in, even a mostly static one like this.

However when composing the camera angles for my screenshots and cinematic I generally just try to make sure my shots are balanced and feel good to look at. The old rule of thirds is pretty much enough for me, while trying to avoid the temptation of too many "security cam angles" meaning cameras placed in the upper corners looking down at the scene. I remember this being all the rage back in my student days, they can obviously work when set-up right but are usually done solely to show as much of the scene as possible. Ending with a poorly composed shot that is commonly just noisy and dull to look at.

Texture Creation With ZBrush

Okay, so, I've been asked to run through my texture creation with ZBrush, it's nothing fancy. Mostly just alphas I made using height map information from scans, basically the same way I make all of my assets. LP base which I then make a HP version of, bring into ZBrush, sculpt, decimate, bring back into Maya (other 3D packages are available I hear) Either retopologize or re-adjust the LP, unwrap, throw into Marmoset (Marmoset's baking tools have changed my life FYI) Bake all the maps I need and then texture with Substance Painter. The "trickiest" step along the way is making sure everything will tile seamlessly but a combination of ZBrush's "wrap mode" and general symmetry takes care of that.

Now don't get me wrong, I love Substance designer and all that it can do but sculpting the base for all my textures brings all the elements together visually more than if I were to try and match my "style" procedurally. Firstly I don't think I have the skill with SD for that but mostly I like the idea that everything in the world was touched by an artist (great name for a crime drama: "Touched by an Artist, starring Benedict Cumberbatch") This way everything used in the scene will also have the same fidelity throughout, by that I mean that the chips, cracks, dents and all those details will have a unified "weight" to them, kind of like texel density but for details.

In Pursuit Of Perfection

The question posed to me here was how I stopped myself from pursuing perfection particularly with remaking an older project. Honestly the answer would be a combination of time and attention span, I'm not sure how long I have to keep doing this art thing and I have so many things I still want to achieve with it. So I get as close to the goal of , "yeah, that's pretty good, I'm pleased with the way that came-out" as I can knowing that my fire for this project isn't perpetual. It's like the amount of coal you have to feed a steam train (Where am I going with this metaphor? This wasn't planned), I have a big pile of coal, don't know exactly how big, but I can make an educated assumption and I have to get as far as I can with it. To continue this poorly thought-out idea, when I came across something that is more difficult or tedious to do I'll have to spend more energy on it, potentially leaving less for the rest of the project. More coal needed to climb the hill, but by taking a break from the piece I am working on or the project all together (Possibly in the train's dining carriage,turns out this metaphorical train of mine now has a metaphorical dining carriage) I can jump to a different asset or aspect of the project to recharge a bit and build-up the motivation (or steam, help I've gone too far with this) to plow through the rest of it. At the end though I'm never 100% happy with a project and inevitably I've learnt so much more since I first started it that all I see now are errors and mistakes. But it reaches a time when you just have to let it go out into the world to fend for itself so you can move on to something else. Who knows, maybe I'll come back to re-do this project in another 3 years and see how rubbish it is by then.

Inspiration and Feedback

I put these two together because they can often work together. Mainly I get inspiration from other artists and what I see around me, I was super fortunate to be working in Paris at the time I started thinking about remaking this project and there is endless inspiration in that city. In particular though the Musée de l'Armée was a great source of inspiration and also reference. I have never been up-close to a real set of field-plate armour before so having so many sets to photograph was awesome, not to mention the weapons and shields. Sometimes though the inspiration for an idea comes from weirder places altogether, I subscribe to a bunch of medieval history YouTubers (I know right, shocking) and while watching a video about the process of forging a sword it was mentioned that Pope Clement the First is the patron saint of blacksmiths. So I found a medieval illustration of him and integrated it into a small painting on the wall, a little bit of extra storytelling from a throw away mention in a YouTube video but there you have it.

Another, very powerful, source of inspiration is the work done by other artists. There are so many talented and skilled people to look up to in the 3D art community that it's sometimes difficult to see their amazing work and not just give up on the thing you are working on but pushing through that and using them as an example of what's possible is the only way to go. In relation to this is the question of feedback. Where do I look for feedback? There are so many places to go to these days like Discord servers, Blogs, ArtStation. All of which I use to a certain degree but I'm lucky enough to be in a group with a gang of great artists where we post our work for feedback. Posting wips on a discord group is super helpful but sometimes your work can get lost in the sea of other posts or you're afraid you're spamming the wip section with constant changes and tweaking. With a more focused group I can post a bunch of images and I always get great feedback points from everyone. Seriously you should see some of the rubbish I post in there, but with all their help it looks so much better than what it would've without their guidance. I really wish I could've told my past self how important it is to seek the feedback of others, not only ask for it but also not be defensive over your work (Which can be difficult sometimes for sure) I'm just fortunate to have a group whose opinions I trust and where ego isn't a factor, when they bring something up it's always going to be for a valid reason. This being said though the various Discord groups out there are a great resource for feedback and I'd recommend them to anyone. Joining challenges is also a good way to go, you should see the first Artstation challenge I took part in...

Oh and also my wife helps a bunch, it turns out that having someone who isn't in the industry critique your work is great because they aren't focused on technicalities, just on visuals. If something looks off she'll say so, doesn't matter how long it took or how difficult it was to achieve. Which is great because that's exactly what the normal player will be doing. It just so happens that not everyone "plays" games by running around looking at the various types of wood that was used and wondering how they got it to look this good. That's just me apparently.

Where To Find Me

(Or where to know to avoid me) Currently I'm working on Chivalry 2 with Torn Banner Studios here in Toronto (Free advertising! yay!) and in terms of personal projects, I have some ideas milling about as to what I'm going to work on next. We all saw the reveal of UE5 so maybe something with Niagara or a heavy use of scans, I'll have to see which idea has the biggest pile of "coal" for me.

Tips and Advice

It feels odd for me to be giving "tips" I mean I still feel like a complete novice sometimes but here we go: The best piece of advice I can give would again be to develop the skill of accepting and asking for feedback, don't be overly precious about your work and I guess stay humble. You'll learn more by listening than by parroting what you know, even if you already have the information often something new and useful will pop up. (Oh and most importantly, stay away from steam train metaphors... they're just the worst.)