Diablo 3

Behind the Gear: Philosophies on making Diablo heroes

diablo8 - Behind the Gear: Philosophies on making Diablo heroes


I wanted to push a little deeper into art methodology and talk about decisions that went into the models for the main heroes. What we made and how it looked was informed by how we made the characters. This post will be a little more technical than some others, but I know that some people like learning about process.

I found a document I wrote, back before Reaper, speaking broadly about how Diablo heroes were made. Here is the first part. Keep in mind, this is specifically about playable Heroes. Monsters, Bosses, and NPCs are made entirely differently. I would add that this list reflects the lessons of an art producer. An art director, being a person with artistic talent and education, would have a lot to say about art itself, things like stylization, technique, and tone.

Philosophies on making Diablo heroes

  • · Create a diverse set of heroes by exploiting differing body shapes, animations, and armor themes.
  • · Maximize character identity with gender dimorphisms- age, size, pose, etc.
  • · Create collections of armor pieces that compile a unified appearance without undermining the mix-and-match realities of an RPG.
  • · Create a progression of 18 complete appearances per class that honors and explores the themes of that class.
  • · Focus the customization of hero appearance in the armor Dye system.
  • · Exploit the fact of our game camera with artistic decisions that emphasize the more regularly-seen head, shoulders, and back.

So, let’s break what those mean.

1. Create a diverse set of heroes by exploiting differing body shapes, animations, and armor themes.

This was a significant decision made very early on in Diablo III development. Hero body differences seem kind of obvious and not a big deal, but they really are.

When you commit to a lineup of heroes with highly diverse bodies, you are committing to a lot more effort modeling, rigging, animation (that’s a big one), and texturing than you would with intentionally similar heroes. The alternative decision is to save time with shared body elements. You can save a TON of work by using the same animation rig on your characters- we did, on many NPCs. In doing using shared rigs, you accept that the characters will have similar animation skeletons, proportions, and have a “samey-ness” about them. That isn’t always bad! In many of the games you play, you are seeing a ton of characters all sharing the same animation rig and utilizing the same or similar animations.

Developers commit to diverse or unique bodies when they intend to make the difference between playable characters one of the draws of the game. Zangief and Charlie look and move nothing alike. Can you image trying to create those characters without giving them different skeletons and animation rigs? They need different underlying structures to do their different things in gameplay. It’s the same with Diablo III- if we tried to make the Witchdoctor and the Wizard on the same animation skeleton, they would come out looking a lot more alike, and would lose a lot of personality.

The last part about armor may not seem like it goes with the others. Armor is just surface textures and snap-on elements, right? Well, it is, but if you commit to having a lineup of truly unique heroes, then the same breastplate armor will look very different on each of them. Most likely, the breastplate will be stretched and smooshed differently over each body, and look bad on all of them. Committing to a lineup of truly diverse hero bodies means a commitment to customizing each piece of gear for every hero. Is the scope of the job starting to sink in?

2. Maximize character identity with gender dimorphisms- age, size, pose, etc.

Back in 2005, the conversation about the representation of player characters was different than it is today. Now, players rightly expect that a game with any hero customization will give you more than one body to choose from. Back then, you could still make an argument for a lineup of unique heroes each tied to one gender. It could be argued that the D2 Amazon is more iconic because there was only one person visualized in that character. But I think we all wanted to make male and female versions of each class. And we knew- the cost of making an alternate gender of each hero is significant. Do not let any armchair game developer tell you otherwise. The time to animate a Diablo hero, with all their abilities and with all their weapon poses, is measured in fractions of a year. Large fractions. Animating a second version of that character takes less time than the first, but still takes months.

In the end, I don’t think anyone on our team regretted the final decision, because we knew what we were buying. We were buying twice the heroes. The male Witch Doctor is one person, and the female Witch Doctor is a very different person. Their bodies are completely different, their ages, their poses. Though their dialog is the same, their voices and personalities are not. Even when the two characters wear the same gear, they look different. That difference is important because when a player sits down to choose their first hero it gives us twice as many opportunities to make a connection with a player. And that is worth quite a lot.

3. Create collections of armor pieces that compile a unified appearance without undermining the mix-and-match realities of an RPG.

This goal speaks to two ideas. One, this was to be an RPG with extensive look customization, and we needed to support that. Two. We wanted to minimize the inevitable patchwork clown appearance of our characters.

It’s a fact of life- if you make a game with character customization of any kind, people are going to hop online with a video of their character posing in pink, blue, yellow, and probably no pants. The unspoken joke is, “look at these stupid developers”, when in reality, they are looking at a testament to player freedom.


The alternative to hideous player freedom is player restriction. Have you ever played a game where you could kind of change up your look, but not the way you wanted? Those devs did choose to commit to the freedom of players to make bad choices. It’s a decision that needs to be made on a game-by-game basis, and in Diablo III, we decided to let players put together many kinds of looks while only adding touches of restrictions in the interest of making heroes look good.

What this means is that we tried to make the armors diverse, out of many different colors and styles, but also steer their palettes and designs into similar patterns. There is no better example of this effort than with the Dye system.

3. Create a progression of 18 complete appearances per class that honors and explores the themes of that class.

I was a part of an effort to create a kind of visual journey for each hero within their armor designs. In the course of gameplay, we knew our players would encounter random loot drops in a generally predictable manner. Low-level gloves show up before high-level gloves. And given that experience, we wanted players to see their heroes decked out in gear that told a visual story of progression. The progression from weak to strong, but also from obscurity to glory. There was a great deal of thought and effort put into this “Armor Story”, and I’ll share more about it soon.

4. Focus the customization of hero appearance in the armor Dye system.

Before the development of the transmog system, we wanted to give players more customization tools than just armor selection. What kind of choices are you making when the best-statted items all look alike? The Dye system, driven by Lead Character Artist Paul Warzecha, was a massive effort to put that control in the players’ hands.

The Dye system, which we internally called Armor Tinting, involved applying tints to specially masked sections of our armor textures. That masking, by the way, had to be added to every single piece of armor in the game, by hand. Remember how all the heroes have different bodies, including the gendered versions? Well, every single piece of armor in the game, times all the classes, needed to be specially modified. How many pieces of armor is that?

  • · Six or so armor pieces visible on the hero.
  • · Multiplied by 18 look tiers, totaling around 100 items
  • · Times 5 heroes (later 7!), to around 500 items.
  • · Times 2 gender versions, to around 1000 items.

That’s a lot of armor files being opened up, masked, and cleaned… like, several months of cleanup just to implement the feature. I remember when we first got the system working on a single barbarian helmet. It looked pretty good, and we were happy with how it worked! I also remember weeks into the masking project, seeing the artists go back over armor sets they had designed a year before and trying to figure out the most logical parts to make dye-able. Internal enthusiasm for dyes took a hit before we were through.

Paul also managed the selection of the Dye colors, which was a very tricky and subjective thing. To feel like actual customization, we needed to provide all the basic color options a player might want. But we wanted to minimize the Fuschia+Cyan+Mint experience. Paul spent a lot of time selecting a range of hues that struck that balance between providing our players artistic freedom and providing the tools of intentional hideousness.

At one point, the team at-large had started to make fun of the dye system while under development. The issue was that we had been testing the system internally by making all new armor drop already pre-dyed in a random color. Everyone had a special “remove all dyes” command to reset this. In theory, this would let the team see what dyed armor pieces looked like without going through the process of locating or buying dye objects in-game.

The result was that everyone’s gear looked ridiculous. The dyes were being selected truly randomly. In our original vision, colors like Brown, Tan, Forest Green, and Charcoal would be most commonly selected, so that a “random” dyed armor item wouldn’t be too far off from what people were already wearing. But these actually random-colored items were giving people clothing in colors like Lemon and Magenta. We got it right soon enough, and the system ended up a success internally and externally. But my heart still hurts when I remember the artists who had spent months making the feature having their work laughed at.

5. Exploit the fact of our game camera with artistic decisions that emphasize the more regularly-seen head, shoulders, and back.

This last one is kind of a no-brainer for veteran artists, but it takes some explanation for everyone else.

When it comes to isometric games, the human body gets treated unfairly. Seen from overhead, the parts of the body that enjoy the most screentime are the head, shoulders, back, and arms. Unless the characters maintain an especially upright posture, you won’t usually be looking at their chests. This means that skin, clothing, and armor designs on the legs, chests, and feet aren’t going to be seen much.

Turning these facts to our advantage took artistic experience and strategy. Our artists created specific areas of low and high detail on the characters, painted light to fall off vertically, and used many other techniques beyond my understanding. As an art producer, I sat in on perhaps hundreds of art reviews, and the depth of the craft still astonishes me. All I can say is that our artists understood that the game camera provided their characters with a very specific challenge, and they worked wonders to meet it.

These philosophical points are kind of a departure from specific gear stories, but some of these details add context to the hows and whys of our hero items. Understanding why we made the heroes’ bodies so different from one another comes into play when we start talking about things like legendary chest armors, and how they were such a tremendous pain in the ass.

Source: Original link

© Post "Behind the Gear: Philosophies on making Diablo heroes" for game Diablo 3.

Top 10 Most Anticipated Video Games of 2020

2020 will have something to satisfy classic and modern gamers alike. To be eligible for the list, the game must be confirmed for 2020, or there should be good reason to expect its release in that year. Therefore, upcoming games with a mere announcement and no discernible release date will not be included.

Top 15 NEW Games of 2020 [FIRST HALF]

2020 has a ton to look forward to...in the video gaming world. Here are fifteen games we're looking forward to in the first half of 2020.

You Might Also Like

Leave a Reply

Your email address will not be published. Required fields are marked *