Polycount/Riot Art Contest

Riot Art Contest – Arctic Fox Ahri
by brdiefenbach
on Sketchfab

 

For the past month or so I’ve been participating in the Polycount / Riot Fan Art Contest. I’ve never been a big League of Legends player, but as a character artist I really dig the game for having so many interesting characters. As with the older VG Remix Contest, I knew participating would be a good chance to learn and grow as an artist so I took a hiatus from my other projects to really dive in and try to meet some personal goals.

But before I go too far, here are my final results as images for those that are unable to view the model in 3D for whatever reason.

 

Project Goals

  1. Use Mostly ZBrush
  2. Shadow of the Colossus Fur Technique
  3. Blender for 3D Painting
  4. Other Lessons Learned
    1. Have a Clearer Concept
    2. Consider the Game View
    3. Provide Critique for Others

1. Use Mostly ZBrush

I bought a personal license for ZBrush about a month before the contest started and I only piddled around here and there up to the time when the contest was announced. I really wanted to dive in and get a feel for how it works and to hopefully be able to use it to take my art to the next level.

So was I successful? Sort of.

I actually kinda felt like this mesh–like I was missing my head and hands when I first started using ZBrush.

I feel like I have a much better understanding of where things are in the user interface. I now know where most of the equivalent tools that I’d use for poly-modeling in Blender are located in ZBrush.

And about 20 days later, we have something not horribly unreasonable as a base. (Note I did not work on this for 20 days straight, but it’s nice to know time stamps.)

I still feel uncomfortable with my technique. Which tools are best to create the shape I want? Should I use this Matcap or another? These are all things I’m still working my way through and will continue to struggle with on my next project.

My understanding of the different “types” of modeling you can do in ZBrush is also limited. I mostly worked with a combination of Dynamesh and ZRemesher, but I know there are other ways to get off the ground and I’m still hazy on the advantages and disadvantages of each method.

My last sculpt WIP update.

I also still have issues with ZBrush as a program. I cannot open a ZProject I have previously saved. Even one that is the default scene with nothing changed! Tools save fine, which I guess is the most important, but it would be more convenient if I could get back some of the other settings. (Quick saves do keep all that information and load up which is serviceable, if not quite ideal.).

2. Shadow of the Colossus Fur Technique

I had seen an article on the SoTC fur technique ages ago and always wanted to give it a try. Unfortunately there wasn’t anything to test it on. As it turns out, Arctic Foxes are ridiculously fluffy, so now was the perfect chance!

405015
Such fluff.

As you can see, I didn’t end up doing this. So what happened? Although I never posted in the thread about this goal, I did some major testing in Unity. Once I got it up and running, I realized this trick just doesn’t fit in the League of Legends style so I gave up pursuing it. Maybe there’s still a way to make it conform to the style, but I didn’t have time to figure it out. So this goal will have to wait for a little longer.

3. Blender for 3D Painting

Back for the VG Remix Contest I downloaded a trial of 3D Coat to try out painting on model. Although it had some nice features, I still couldn’t quite justify purchasing it as Photoshop and Blender did basically everything I needed. (Blender can’t import PSD files like 3D Coat, but it can still help me paint over seams as a last step). As I’ve gotten more into hand painting, I’ve wanted to be able to paint in 3D once again.

Luckily, the latest version of Blender introduced new features to their 3D painting toolset. It’s still not Photoshop, but it’s much better than previous versions. This video tutorial by XRG81 really helped me get a better handle on how to approach painting in Blender (and how to make it feel a bit more like Photoshop).

 

I also took this as an opportunity to try out a grayscale to color approach. Once my terrible-because-of-my-sub-par-sculpting bakes were made I was ready to paint! Alternating between Blender and Photoshop was pretty smooth (saving from Blender as TGAs for best results). Once I felt the grayscale values were good, I used gradient maps to quickly adjust the colors and then get back in to clean up any weirdness that was left.

4A. Have a Clearer Concept

Above you see my concept art for what I wanted the character to look like. As you probably can tell, it was a quick and dirty exercise. Through some quick Google-Fu I had some reference images for Wildings in Game of Thrones, Traditional Arctic Clothing, official and fan interpretations of Ahri, and photos of Arctic Foxes. Although all the resources were helpful, I didn’t end up committing to a particular vision.

As I did some research I realized there were several types of fur: you’ve got your fashion style fluff-to-the-max, the more rugged GoT fur, and stuff in between. At first I was really leaning towards the bulky fashion fur, but as I struggled to make shapes my concept watered down into a more smooth material.

I tried to reconcile the two ideas in the last stages of my sculpt by adding in some of the rugged GoT fur along the coat-tails but when it came down to baking the details didn’t transfer. As I went on I was forced to give up on the fur look for the main part of her clothes entirely.

There were things about my concept that I did genuinely like. I wanted to balance functionality in a cold climate with Ahri’s sexy character. I also really wanted to maintain her silhouette since those are pretty important in the MOBA genre. In these aspects I think i did a great job and the changes I made to my concept as I progressed really help keep a balance between her old design and my new skin.

Some of those changes included:

  • Changing leggings into leather boots
  • Changing tiny runes into bigger patterns
  • Taking the hood off to allow her ears to be more prominent

4B. Lessons Learned – Consider the Game View

ahri_texture_03_notes

The above points sort of lead into the next one: Thinking about the game view, or maybe I should say designing your character for the game view. When I got to the later part of my texturing process I realized that there were some areas of my UVs that I wanted to change based on what you could see well from the game view.

The areas circled in red are really visible from the high-angle camera. But they don’t take up that much of my texture. I guess it’s about a quarter, which is a decent chunk. But consider the back of Ahri’s coat. That area, is covered by her tails almost all the time. If I had thought more closely, I could’ve spent less time sculpting it and put that time into these other areas in addition to giving them bigger texture islands.

Another thing was small details like my original runes and the lines on Ahri’s face. They look good in an illustration, but when you look from the game they were just noise that didn’t help really convey anything important to a player.

Related to this is that there are some mesh optimizations I probably could’ve gotten away with because of the camera angle. Diablo 3 has a really poignant example of this, where the devs turned the trees in the levels into billboards because it was easier to paint trees and cost less computational resources. In my case I have geometry in several places that you probably wouldn’t see from the game camera or if you did a double-sided plane would have worked equally well.

4C. Lessons Learned – Provide Better Critique for Others

One of the problems I had during this competition was that no one would provide any feedback! Early on when I was still hammering in proportions I received more comments and links to reference images and such. But once I had basic anatomy down, I stopped receiving comments on my character model.

This is one of those frustrating things that isn’t really limited to the contest, but seems to be an issue with every art forum (and even in person). You’ll only get help if:

  1. You’re terrible.
  2. You’re amazing.*

In the first case, people will point you toward reference images. They might point out mistakes and provide paint overs and generally encourage you. In the second case, you’ll probably get a lot of attention and folks saying “Looks amazing” or “Great job!” or “Can’t wait for more!” But you might not get the level of feedback you’re really looking for.

I realize now that the competition is close to over, that I’m also guilty of not providing the type of feedback I really wanted. I spent most of my time posting about me and not looking around at what other folks were doing.

So for next time, I guess I’m just going to try to make more of an effort to actually critique things. To do more than say great or awesome. But if something is looking great or awesome, I want to reassure the artist that they’re doing the right thing and to keep it up!

 

Well that’s all folks! If you actually read this far, congratulations!