What do the graphic designers have to say about creating the Colossi and ancient lands that drew players into the world of the game? The staff discussed this in an extended interview found in the Shadow of the Colossus official guidebook, which has been translated below.
Creating, Testing, and Tuning the Models: A Cycle of Repetition
Design Sketches from Shadow of the Colossus
Were the characters that appear at the beginning of Shadow of the Colossus the first to be created?
Ueda: The human characters were the first to take shape, but the Colossi were the first to be completed. I was always the one responsible for the human characters, but I asked the staff to handle turning the Colossi into 3D models.
Niwa: We’d get ideas from Ueda and expand on them ourselves. Questions would arise about what kind of arena would be necessary based on what form a Colossus would take, so the character design team and the field design team began working together.
Suzuki: We didn’t have any particular design in mind at first, the game sort of came to be the way it is today through our fumbling efforts…
Ueda: The design of the Colossi is closely related to how you go about defeating them. That meant that even we were successful in designing them, we wouldn’t know whether they’d work in-game or not before actually modeling and testing them out.
Niwa: When things didn’t go well during testing, the Colossus’ design would be changed and we’d end up having to alter the arena as well. That meant there were frequent exchanges with the field design team. A really small error on our part would sometimes create huge problems elsewhere.
In what order did you design things?
Suzuki: We’d start by drawing up a simple document that detailed the shape of the Colossus and where fur would be growing on its body that the player could use to climb up. We also wrote down how the battle would flow and information about the Colossus’ arena. There was never any rule that said we could only use a particular design. All Ueda told us was the size of the Colossus and the pixel resolution of the screen, so as long as we kept those limits in mind, we were free to do as we pleased.
Niwa: Looking at designs from early in development, there’s one Colossus that’s just a giant bird. The balance between living creatures and inorganic ones is something that we were gradually able to achieve over the course of the game’s development.
Suzuki: When we first started making the game, we didn’t know how far we could take things. We had specifications for Colossi that threw things and Colossi you’d climb after lassoing them with a rope. We’d make bits and pieces, then problems would arise and we’d give up on trying to implement them. We were still trying to fit things in right up until just before the game’s completion.
Ueda: We steadily applied different ideas in order to increase the players’ sense of scale and the pixel resolution of the screen. If the players don’t get an idea of these things, it’s impossible to achieve consistency. The simpler the grip points are, the less problems occur, but the resolution and sense of scale disappear when you look at the game as though it were a painting. It’s difficult to find a balance.
Suzuki: Consequently, we wouldn’t know if something worked until we tried it. This lead to a repeating cycle of creating something, testing it, tuning it, and testing it out once more. Also, players had to be able to defeat a Colossus by using the method we’d decided on. There were times when someone would find another way to defeat it. On other occasions, there would be places you shouldn’t have been able to get to but were able to reach through the movements of a Colossus. We went to a lot of trouble trying to make sure that players wouldn’t be able to reach anywhere they weren’t supposed to be.
What sort of smaller objects did you create?
Niwa: I modeled things like bows, bow guns, and small animals like birds and fruit, in addition to all the times you get in Time Attack mode. I did the smaller animals according to Ueda’s specifications. I wanted to make a cow, though…
Ueda: Even if you had, there’s no way I would’ve put it in. [laugh]
Niwa: The forest once had a lot of birds in it, but they were all cut due to memory limitations, unfortunately.
Expressing the Enormity of the Ancient Lands
Players get a sense of the atmosphere of the game world, but what was it actually like to create?
Hasegawa: I don’t think the modeling process itself differed much from other games, but, in my opinion, we were able to project the atmosphere of Shadow of the Colossus by adding fog and coupling it with a natural landscape that wasn’t particularly wild or untamed. Because the game draws even far-off scenery, we were able to project a comprehensively unique atmosphere.
How were the Ancient Lands created?
Hasegawa: We started out by creating the Colossi’s arenas, then figuring out where in the expansive field they would be placed. There were times when the width of the field was decided by how the arenas would connect. Since we started out with 24 Colossi, it was pretty huge. The number of Colossi was ultimately reduced to 16, however, and the field was reduced a bit to its present form.
It’s still quite large, regardless. It’s incredible how seamless the transition from place to place is, no matter where you go.
Hasegawa: It took a lot of time before it became the way it is today. At first there were no distant-perspective models, and even after we made them you weren’t able to see them unless you got within a few hundred meters. It wasn’t until comparatively late in development that we put in the furthest distant-perspective models. It wasn’t until all the little adjustments started adding up that we began getting feedback about how our system was able to display even vast expanses.
How did you organize the scenery?
Hasegawa: Elaborate polygon models are used within the immediate vicinity of the player. The further away from the player a model is, the less polygons it has. The furthest away models with the fewest polygons mostly include cliffs and the surface of the earth. Even those models’ polygons aren’t drawn completely, they disappear at a certain distance. However, we tried to make it so that areas that command an excellent view are expressed in more detail, while areas that are hard on memory aren’t expressed so elaborately. In addition, usually games have clear barriers to indicate where players aren’t able to go, but we tried not to do that. We wanted to make it so that players can reach anywhere it looks like they’ll be able to get to.
The way light is expressed is also impressive.
Ueda: The world in the game actually does have night as well as day. I’d wanted to have time and weather variations ever since making Ico, but it’s another thing that was cut due to memory issues.
Hasegawa: However, because the direction of the sunlight depends on factors such as the time of day and the weather, we tried to have a sense of overall unity. Strictly speaking, however, it sometimes differs by area, to a certain extent. Most of the time, when I’d submit things for Ueda to check over, he’d tell me “Make it whiter! Increase the brightness and the saturation of the textures!” [laugh} I’d wonder if it wouldn’t be too much, but I went along with what he said. Now, when you reach a place where the sun is shining, you get a sense of dazzling radiance, so in the end it worked out well.
What did you struggle with, Kajita?
Kajita: Only that the amount of work was overwhelming! [laugh] In the latter half of development, it was common for the game’s appearance to change every other day.
Kaido: We had a map that divided up the areas people were in charge with using a colour code, so it felt like you were looking at a simulation game. [laugh] The Kajita Empire occupied the southern territories, for example. It was afterwards that I appreciated just how much he had done. [laugh]
Hasegawa: The data was all compiled in one place, meaning that anybody could mess around with anything. Everyone had basic responsibilities, but Kajita would go all out, fixing up all the textures. [laugh]
Kajita: Ueda’s creations stimulated me. I’d think to myself, “Alright, I have to make something better than this!” and my motivation would increase.
When you were creating the field, were you influenced by any particular region or country?
Hasegawa: Not me, personally. In order to get a realistic sense of scale, though, I used the buildings I saw biking to work each day as a reference.
Kajita: Ueda would suggest something to us and we’d rack our brains for the rest… At first, we didn’t know what Ueda had in mind, but once the middle of development rolled around, we had him pretty well figured out and were bolder as a result.
Was there anything you struggled with concerning the Colossi’s arenas?
Hasegawa: Much like with the creation of the Colossi themselves, we had a repeating cycle of creating, testing, and tuning.
Kajita: Leo’s (#11) area was profoundly memorable for me. The pools of water that remain around the completed arena were previously an entire lake. It wasn’t just that we changed how to defeat it and thus the arena came to be the way it is today, Ueda told us to make it seem as though, long ago, there was once a bigger lake in the area. It was fun to create.
Ueda: Actually, the Colossus was once swimming in the middle of the lake. But when we tried to make his swimming speed match Wander’s, he looked weak, so we gave up on the idea.