Cubemaps and Reflections

    

    Welcome to my second blogpost about our 2nd homework in Computer Graphics 2 class.

    In this assignment we worked with static and dynamic cubemaps and reflections. Again I worked step by step and I will be sharing my process with you.

Figure 0: HW2 final result


First Step: Static Cubemap

    Firstly, I wanted to show our environment correctly before going to other steps. This is done with static cubemapping. After creating a cubemap texture from the given 6 images, I put these textures on a cube object with [-1,1] dimensions to work correctly with texture coordinates. After putting my camera on (0,0,0) position, I was able to see my environment. 


Figure 1: Static cubemap result

Second Step: Ground Texture

    After showing environment, I wanted to show the ground. I used the ground object that was given to me. But in addition I assigned it texture coordinates given below, to create a repeated texture on the ground rather than having a distorted result. I also rotated the ground object 90 degrees to have a horizontal surface.


GLfloat groundTexCoords[] = {
0.0f, 40.0f,
40.0f, 40.0f,
40.0f, 0.0f,
0.0f, 0.0f
};


    To see the ground correctly, I configured the eye position and look at positions when looking at the ground to be same with the car's camera.

Figure 2: Cubemap + ground

Third Step: Rendering the Car and Camera Positions

    After ground, I wanted to show our car and have the correct camera positions that depends on the user input. On default, user sees the car from behind. Firstly, depending on which side the user wants to see the car that is on the (0,0,0) position initially, I rotated the camera positions and look at positions by 90, 180 or 270 degrees to show the correct side of the car. I also applied this rotations to the static cubemap.
    Secondly, I implemented the car's rotations on user input. I rotated the car and its camera by the current angle that increases and decreases according to user input around the y axis. I also applied this rotation to the static cubemap.
    Last step was the car's displacement. Since user changes the velocity and not the car's positions directly, I kept the velocity. Normally without any rotations the car goes on the (0,0,1) vector. By rotating this vector according to rotation by user input and multiplying the result by the current velocity and adding this to the car's last position, we get the new position. For the camera, I translated it by the car's current position.  A side note here is that I did not apply this translation to the cubemap camera to have the image of a infinite environment.


    glm::mat4 carRotMat = glm::rotate(glm::mat4(1.f),glm::radians(carRotAngle), glm::vec3(0,1,0));
glm::vec4 carX = carRotMat * glm::vec4(0,0,1,0);
carPos = carPos + glm::vec3(velocity * carX);

n

Figure 3: Rendering the carr


Last Step: Statues, Dynamic Cubemap and Reflections

    Lastly, I needed to make the car reflect its environement.
    Firstly, I added 8 bunny statues around the car in 45 degrees all in different pastel colors. These statues
remain still during the whole game.
    Then I had to create a dynamic cubemap so that we can have the correct reflections on the car in every
frame. To do this, I created a color cubemap texture, a framebuffer and a depth renderbuffer. I binded the
texture and renderbuffer to the framebuffer. By looking at the 6 directions from the car's position, I was able
create the dynamic cubemap.
    A little difficulty I had here was that firstly I forgot to add the depth buffer to my framebuffer so my statues
would have a distorted reflection. After adding the depth buffer I was able to get clear reflections.
    Another little note here is that while creating the cubemap, I looked at the statues and the ground from the
car's position but to the static environment, I again looked at from (0,0,0) positions since we don't want to include translations
in the environment mapping.


Figure 4: Looking at the dynamic cubemap texture
(inside the car)


Figure 5: Final image from left

Figure 6: Final image from right




Comments

Post a Comment

Popular posts from this blog

Surface Rendering with Bezier Surfaces

Image
Welcome to my first blog post which will be about our first assignment in CENG469.      In this assignment we were asked render the surface whose information was given in the input with ambient lighting and diffuse and specular shading effects of the given point lights. We also needed to be able to change the sample size, rotation angle and coordinates of this surface according to user interaction. First step: Render a flat surface      Firstly, I wanted to be able render a flat surface without any height component on the XY plane. First step was parsing the input and calculating how many vertical and horizontal bezier surfaces there are. Instead of sending all of the vertices to vertex shader at once, I have created sampleSize x sampleSize vertex buffer and created its index array buffer. Then for every bezier surface, I sent this buffer to vertex shader with the horizontal and vertical Bezier surface counts and the surface’s index. Using these data, verte...
Read more

My Snow Simulation

Image
 Welcome to my last blog for our course CENG469 Computer Graphics 2! This blog will be about our final project which we chose the subject ourselves.  I chose to simulate a falling snow and its accumulation on ground as I am in the end a winter child and also wanted to create something I encounter while playing my favourite games. I implemented a particle system to simulate the falling snow and used displacement map to show the accumulation. I will again explain my implementation step by step. Snow Particles First step was generating the snow particles. To do this I implemented a snowflake struct which has the necessary attributes position, size, velocity and direction vector. In every frame, I created a new snowflake with random attributes and erased them when they hit the ground. Also I added and normalized their direction vector with another random vector to create the chaotic movement of snow fall. struct Snowflake { Snowflake ( glm :: vec3 p , glm :: vec3 d , float ...
Read more

Quaternions and Procedural Noise

Image
 Welcome to my third blog post which will about our third assignment in Computer Graphics 2 class. In this assignment, we were asked to implement a plane-like movement camera in an environment which has cloud texture made with Perlin Noise. Transformation with quaternions I first started with implementing the plane like movements which are yaw, pitch and roll of the camera using quaternions. To capture the movements, I first created three vectors which represents the axises of the environment. Then, everytime the user presses a button to change the rotation, I calculated the related quaternion and applied it on these axises. Then I did all my other calculations (such as placing the quad for cloud texture which will be mentioned in a bit) according to those axises. glm :: mat4 rotationmatrix = glm :: toMat4 ( rollQuat ) * glm :: toMat4 ( pitchQuat ) * glm :: toMat4 ( yawQuat ); xAxis = glm :: normalize ( rotationmatrix * xAxis ); yAxis = glm :: normalize ( rotationmatrix ...
Read more