3D models are getting more and more realistic as they become more accessible to users.
According to new research by a group at Northwestern University, models in 3D games are better at conveying physical information and that they also better simulate the interaction between human bodies.
The research shows that models can help the player, but the data shows that the games also require additional skills.
“In a typical simulation, there’s a single, central computer,” said J.J. Johnson, the lead author of the study and a senior researcher at Northwestern’s Kinetic Computing Lab.
“But in a 3D game, there are a lot of computers all around you, and each of them has to learn to interact with the rest of the environment in a way that makes it feel realistic.”
Johnson and his colleagues tested more than 200 3D simulations using two different approaches: using a game engine that uses a 3-D camera, and a program called a 3d renderer that uses data collected by the human eye.
The researchers found that the programs had significant advantages.
In a game using the game engine, the players’ body was represented with a single object in the world.
In a renderer, each object in a scene was rendered in three dimensions, so that it was possible to see the difference between the height of an object in space and the height at which that object was visible in the camera’s field of view.
Johnson said the games that used the renderer outperformed those that used a game camera in many ways.
“We think that a lot more players are going to be interested in the 3D renderer in the future,” Johnson said.
“The game that uses the renderers is going to become more popular.”
Johnson said that in order to simulate a swimmer’s body in a realistic manner, the renders must be able to do a good job of simulating the movements of the body.
“So we wanted to see if we could simulate the body movement of a swimsuit model that’s really high-quality,” Johnson explained.
The model’s head was modeled in the rendered model.
The body’s position was simulated by using the rendering program.
The human body is represented in a single mesh that is modeled in 3 dimensions.
In a rendered renderer the position of the head was simulated.
In another renderer and with a different model, the position was replicated using a single model.
The same model with different mesh sizes and a different renderer was used.
“The body has to move around a lot, and the model has to do some work to figure out where it’s going to end up,” Johnson added.
“In a 3DS [3D stereoscopic display] it can look really good.”
The researchers used a variety of models from models made by Microsoft and Sony, including a swimwear model from the popular Sony PlayStation 3 game The Swapper, and from a swimsuits model made by the manufacturer Pinnacle Entertainment.
“There’s a lot going on in the body that needs to be modeled,” Johnson noted.
“It’s very hard to see that in a renderers and a 3ds.
It’s a really challenging task to have a model that has these very complex movements.
But the renderers do it.”
In a third simulation, Johnson and his team simulated a swim wearing a bikini.
They found that in each of the three simulations, the model had a lot to work with.
“It’s pretty easy to understand what the movement of the model is, and what it’s doing to the body in the scene,” Johnson admitted.
But the models needed to be able be scaled to work in 3ds and 3d.
Johnson said they found that these renderer programs did a good enough job to allow for 3D body movements in a game.
“This is a really, really good renderer,” he said.
“I think this is a great place to start to go and get more sophisticated models, and we’re going to see a lot [of] 3D-based games going forward.”