3D Environments, Atmosphere, and PvM Combat

You probably don’t know this, but we started WAVEN (which at the time was called KROSMOZ WORLD, then DOFUS CUBE, and then very briefly WAKFU HEROES) about two years ago now. There were only two or three of us working on it, but the idea was to figure out what we wanted and what our goals were.
We started by thinking about a zone in order to define a graphic style. This zone took the delirious form of something we dubbed “Arachnee-Man’s Tower”. Take a look:




There were already some atmospheres we liked a lot, but Arnaud and his team kept reworking the details, adapting to the game design that was gradually taking shape. The major difference between this version and what we’re currently developing has to do with the game design. Krosmoz World had a few ideas that we’ve kept, but the character classes and spell acquisition was much closer to our current MMOs. Not to mention we certainly hadn’t incorporated the idea of a cross-platform game. In the end, we went with a more “cubic” rendering, and larger cells for combat.

Below, you can see the progress we’ve made on our scenery for the city of Bonta. As you can see, the houses, surfaces, and other “props” are gradually being finetuned. Overall, the renders are close to our background artists’ early concept art, but emphasis has been placed on the “cubic” aspect to enhance the link between the open world and combat. In WAVEN, we don’t want to depend on a grid and want the combat fields to be readable.



WAVEN and 3D Scenery

Now that you have a sense of how we’ve progressed in finetuning the scenery, I’d like to talk more about the topic of this post and our decision to go with 3D environments. The devs clearly explained that developing the scenery in 3D would give us many advantages, which I’ll list below.

  • 3D in UNITY: For this highly technical section, I’ve called on a development wizard, Harry Codder. Harry, go ahead and explain the advantage of 3D in Unity.

    “By definition, 2D has no volume and therefore cannot spread light or project coherent shadows in a 3D world. If you want to change the color of the light or where it’s coming from in your scene, you have to re-draw everything! If we wanted dynamic scenery and atmosphere, it was going to be nearly impossible for us to continue producing them in 2D. As far as advantages, 3D scenery elements can be re-used from one environment to another, and they can face different directions without having to draw a specific version each time. This makes level building much easier, and testing different versions of the same zone is much faster.
    The ability to move scenery elements without having to create a traditional animation specific to the movement makes it possible to create many more animations, thereby making the world more interactive. From a purely technical standpoint, 3D isometric rendering in 2D comes with many constraints in terms of setting up an effective rendering pipeline. Managing the element rendering order alone is a puzzle that is easily shattered by the latest brilliant idea by a somewhat ambitious game designer. (Tot: Don’t get snippy, Harry, or I’ll call he-who-must-not-be-named.) Producing environments in 3D lifts these constraints, and we can take rendering further than before with the use of effects that are impossible to achieve in two dimensions.”

    A huge thanks to Harry Codder for his contribution. It’s a major first, and I hope he’ll agree to come back (even if he did somewhat spoil a few things I’ll discuss below…).

  • Re-Using Assets and the LEGO Factor: In DOFUS and WAKFU, if we want to duplicate the same house and just change the door or window to build “rhythm”, we need to create different assets systematically. With 3D, however, we can pool the various parts of a house to compose and re-compose houses however we like. It’s more streamlined and less chaotic for level designers.


  • Lighting, Dynamic Shadows, Reflections, and Other Weather Effects: After a few tests, we saw a clear advantage of 3D in this area… But rather than give a long speech, I’ll show you an example below. But remember our deal, dear readers: You know that EVERYTHING I show you could change, and these are STILL tests and WIP versions. Here you’ll see a few NPCs in action, and especially the type of atmospheric rendering we’re working towards. Also keep in mind that this is only a video, but this example was taken from the game engine and the events will be random. What’s more, I must have lost an hour asking Seb to re-play the scene to see the various possible versions…

    As you can see in this example, it all works very well and you can’t tell it’s 3D. On the contrary, that’s exactly the point.
    I’m presenting this topic because we have absolutely decided that the scenery will all be in 3D. That may seem insignificant, but it’s a first for us. There was one firm condition for this: 2D and 3D environments had to be visually seamless. If you look at the images on the top, which were done in Flash, and the videos, I think we managed it well.
    Many thanks to Pitch for adding great sound to it, and under such a tight schedule.

  • Starting Combat and Developing PvM Combat Instances: Okay, I know my subtitle is a bit clumsy, but this subject is particularly important to me because it relates to starting fights with monsters. I’ll try to keep it short, because this is a subject that deserves its own post. So, I’ll just give you the gist here. In our earlier games, you trigger combat by clicking on a monster or group of monsters. You always know what you’ll be up against, because it’s “listed”. In WAVEN, because of what we wanted the game design to be, but also especially for immersion, we came up with another system. The idea is to incorporate the monsters into the scenery, hiding them (or not). It’ll be up to you to observe the scenery and start the fight when you want to once you’ve spotted a suspicious group. Below, you can see two examples from Tiko that show what we want to do. Obviously, this is the basic idea, and we still have a lot more to develop in this area. But this will give you an idea of the concept.

    What I like most about this idea is the overall “cartoon” element to it, and the possibilities that opens up for us. Here you have a “simple” example, but we could have Boowolves jump out from behind the bushes, have a huge dragon plop down, etc.

    In pure game design terms, the aim is to have a specific family of monsters per zone, and to have this family “evolve” over time. Discovering archmonsters or other super creatures won’t be random luck, but rather the result of your efforts. Let’s take the Crow Forest, for example. When you go there, you’ll find the first few Crows and defeat them easily. But the more Crows you take out, the more you’ll have access to the next group of Crows, which are bigger, rarer, and more epic.
    The idea is to have a table of accomplishments for each region and offer you fights in line with what you’ve accomplished in the particular zone. Thus, the “farming” aspect of the game is enhanced and motivating, because each fight lets you progress in your zone achievements.

    Regarding the levels of the groups you’ll be attacking, we would like these to be linked to visuals. If you walk past a tree containing a single Crow, you know your fight will be simple and not dangerous. However, a tree containing five or six Crows will indicate a much more dangerous fight.
    And what does 3D have to do with all that, you ask? Well, combat zone creation (unboxing cells) is much more practical this way, and movement on cells is less “performance-heavy” than what’s possible in Krosmaga.


Excerpt from Tot’s blog, November 22, 2017.
Read the original post (in French) in its entirety.

Leave a Reply