The Layered World of ‘Breath of the Wild’ — The Art and Implementation of 3D Graphics |
This presentation was carried out by an artist, Satoru Takizawa and a programmer, Takuhiro Dohta, and they explained how the graphics of BotW were based on the idea of “layers,” and how they were the results of the combined efforts of the artists and programmers.
Takuhiro Dohta |
Satoru Takizawa |
Originally at the beginning of development, each and every plant was displayed in fine, clear detail from corner to corner of the map — but this actually appeared cluttered and messy. That’s why they decided to try and simplify objects the further they appear from the camera. The symbolism could even be cartoony — in fact that would appear more natural. It was even a logical choice given how information is omitted from real-life paintings. Indeed, what they were to do was the perfect compromise between realism and symbolism.
When the programmers heard this they started by experimenting with atmospheric particles out on the field — the way particles floating in the air reflect the sunlight. By implementing this they would be able to introduce complex coloring and transparency. The hazy air filled with moisture after rainfall; the glint of sunlight and the dust that tends to light up from that as it dances in the air. It was their challenge to see how the camera would take in that mix of light in the air. They prioritized what would be the most effective over what would be physically correct. And once again, they managed to punctuate that mix of realism and symbolism. They also used classic fog in tandem with that atmospheric effect — a trick of the trade artists have used for ages to add that finishing touch. They say that this fog was used for places like the sandstorms in the desert and the fog in the Lost Woods. They also took obstruction into consideration, which lead to poignant touches, like the way light appears stronger as it beams through a window, or through the gaps of tree branches. In this way the programmers created, and recreated tools to express the air — the “breath” of the world- and passed it off to the artists.
As a result, artists had over 200 different parameters at their disposal. The programmers anticipated that such a number would be too overwhelming — no one could learn and be effective with all of that — so they whittled down to 50 essentials. In doing that the far-distance could be adjusted with particle effects, the mid-distance Y axis could be adjusted with fog. and up close obstructed particle effects could be used— providing clear differentiation between each method at their disposal. It was this tri-layered structure that would be the focal point of their art. These would be the elements that comprised their art — the elements utilized for drawing this space. In order for players to enjoy running around this world they needed a variety of terrains, culture and various climates. By multiplying these three into various combinations every different part of the world would feel special. There’s one more important concept that hasn’t been discussed yet, however — time. The artists thought it best to express a world that lives and moves so that players would never tire of, rather than one static, but beautiful space.
For an adventure, it’s important that sometimes players experience the kind of nature that they wouldn’t expect. It was vital that the world live and move. Environments should change between morning, day and night. Weather should transition. The sun should blur in a cloudy sky, there should be contrast in wet terrain, snow should be detailed. They also wanted to implement something dramatic — like a red sky showing during special events. The artists illustrated these grand images to hand off to the programmers. Combining time and weather would bear variety. Combining time and space would then multiply the number of different artistic designs possible. The first thing the programmers thought of when given this request was that they would need to implement good, stable code and art that wouldn’t allow any variety of 3D models to break. It was important that regardless of what kind of ideas the artists had to create their space, it would have to feel the way they intended and appear appealing. At the same time, it was also important that the game was stable and ran correctly. Because of this, any rendering methods that would require pre-computation and couldn’t be checked in real-time were out. They needed completely real time rendering, where whatever they wanted to render would show up right away. So they devised various methods that worked in real-time: main light using the sun, depth shadows from obstruction, dynamic cube map shadows, diffuse ambience.
And with that they had all the tools they needed to make changes in real time. They started focusing on what they could play with next — and that’s how they came up with the look for the towers and shrines springing out from nothingness. They had the ability to combine the different weather conditions with any different time AND change the coloring of the space in real time. As a result, they had 1,440 different varieties of “space” that they could give Hyrule. This should give an idea for how the idea of “layering” took shape in all the different tasks that were laid out before them. This goes for the combination of the three different distances, as well as the combination of space and time to create the world. You could even say that there were layers of communication in how the artists would devise ideas and the programmers would come ups with methods to help bring those ideas to life — artists can draw and programmers can code. By each dev expressing themselves using what they do best, they comprised the very highest layers in the structure. And this method itself is what brought the art of the kingdom of Hyrule to life.
[Source: IGN Japan]
No comments:
Post a Comment