Welcome back, everyone!
Today’s post is the second article on the subject and I really recommend, if you haven’t already, you read the previous instalment before you read this one.
Last time we explained how we had to start from scratch when we were developing our new characters, and today we’re going to talk about the more critical and curious technical aspects!
We’ve already mentioned that, in order to start redesigning the faces, we had to redesign the bodies as well, all the while keeping in mind what moves and movements our playable characters had to perform.
So we decided to use the level concept-art we were working on (something we’ll discuss more in detail in our next posts) as a starting point since every level was built following a metric grid. This was the only way we could know exactly how high and far our characters had to jump, and how large an environment had to be for them to be able to run or slide through them.
Their form had to reflect everything we had described in the book, everything we had started to define with the new face designs and everything that was necessary for the game’s playability… a piece of cake, right?!
Well, after dozens of sketches (all the details of the design process will be explained by Gianluca in a future article), we finally found what we were looking for: athletic and well proportioned figures with a slight “super-deformed” effect, the correct proportions for a game rather than a film, and with all the physiological details from the story. This last point really made me bang my head repeatedly against the wall.
As you may remember, an Althanoid’s body is covered with two special materials: Bariocortex and Alluronium. Their torso is mainly made of this second semi-transparent element, which allows us to glimpse the Layer, i.e. their robotic skeleton, thanks to a translucent effect. And here comes the first of our problems: do you know how hard it is to manage a transparent effect in a video-game? We, unfortunately, found out the hard way!
Thanks to a rather complicated and tricky process called “rendering pipeline”, used by every video-game system to “render” (or more simply “show” things on screen), that draws transparent objects last (this isn’t always the case, but let’s keep it simple…). For example, imagine a scene in a room with a table and a glass on it, well, the programme will draw the glass after all the rest of the room and the table have been drawn. But what if you have an object that’s only partially transparent? And keep in mind that both the Alluronium and the Bariocortex have a translucent effect (the same you see when you place the tip of your finger on a torch: the light “shines” through your finger!) so they light up the body from the inside in very specific ways. Argh, that’s another effect!!!
In the end we had to develop a special “shader” to render the Althanoids. A shader is basically the programme you need to tell a video card how to represent every object you see on-screen (and we don’t only mean the transparent ones: every object, 3D and 2D alike, needs a shader so it can be rendered).
To be honest we ended up having to develop two different shaders, influencing how we designed our main characters Rise and Faith: both their torsos and the tops of their arms belong to “separate” models from the rest of their bodies.
In fact we chose to make most of their bodies translucent instead of see-through, whereas the torso is both transparent and translucent. This way only the torso is drawn last, solving the problem of how we had to organize and order our transparent objects.
Was that it? Not even close! Actually our problems continued right until we added the “DOF” (Depth of Field) effect that blurs images in the foreground and background, just like with real cameras. When we applied the DOF to the Althanoids’ bodies they were, more often than not, a lot more blurred than we wanted! Why was that? Because of the transparency! Practically the programme calculates the DOF effect using a special tridimensional map that shows the depth of every scene, so it knows exactly what to blur or not. Pity that transparent objects don’t appear on this map and the DOF didn’t have a clue as to what characters were in the scene!! ARGH!!
After a long period of trial and error, we finally managed to find a way to show the characters on the map so they could be read correctly by the DOF.
Having lived the process from start to finish, I can say without a doubt that, had I known what I was in for from the start, I would have avoided creating characters that needed these effects, but it would have been such a shame, because our creatures are really awesome!
When you play X-Light, I’d like you to take a moment to notice how Rise and Faith’s bodies respond effortlessly to every scenes’ lighting, and tell us if it was worth all the effort.
In our opinion, it totally was! 😉
Till next time,
- 0 Comment
- December 6, 2018