This time we are going to get into a little bit of code and produce the simplest of all 3d solids, the tetrahedron. I know its not the most exciting of things but we have to start somewhere. The scope of 3D graphics in computers is so vast that its very easy to get lost in the vast piles of research papers.
First lets do some basic setup, if you followed my last post then you will will have a project template to use, this makes this a little easier.
For those of you that are running on Windows and want to use Visual Studio please leave a comment if you would like
a project template for F#. The beauty of MonoGame is that it is cross platform and there is only a small amount of code
that differers between the different platforms, and that is localised to the main entry point rather than the
First create a new
MonoGame Mac Application project, you should end up with a Game1 type that looks like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
We are going to need a few extra field for this sample and we wont be using and 2d so remove the
spriteBatch and the
logoTexture as we wont be needing those. The following fields need to be added in their place:
1 2 3 4 5 6
What’s this! mutable fields! I know, but this simplifies things until I can put together a friendly functional
scaffolding around MonoGame. We are creating a
basicEffect, this is used to draw the 3D objects, its actually
just a basic shader implementation with simple lighting. We also have
texture which will be used as our texture
map. We have a
vertexBuffer which is used to store the vertices for out primitive.
are our matrices which are used to look into our 3D scene. For more information on the theory behind 3D projection have
a look here.
Lets more to the
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
The first thing that we do is load the texture map:
texture <- x.Content.Load<Texture2D>("Tetrahedron")
This simply loads in the texture named
Tetrahedron using the content loader.
Next we set up the default values for the world matrix, view and projection matrices. The
world is simply initialised using
Matrix.Identity which is a matrix defined as:
1 2 3 4
view is initialised using the
CreateLookAt method of the Matrix type. This sets up a transformation that points from
0,0,10 to the centre of the world using
Vector3.Zero. It also uses the
Vector3.Up as the orientation direction (Positive Y is up).
projection is also initialised using the helper method
CreatePerspectiveFieldOfView which as you might guess, creates
a perspective with a field of view. In this instance our field of view uses the constant
The basic effect is now initialised with the matrices we just initialised.
For now I want you to ignore the line
let tetrahedronData = generateTetrahedron 3.5f. I need to explain how to generate a
tetrahedron before that will make sense, just assume that is returns the vertices that we need for the tetrahedron.
vertexBuffer is now created which will hold all the vertices for the tetrahedron. We need to tell the
what format we want to use to hold the vertices, here, we are going to use vertices with Position, Colour, and Texture
coordinates so we we use the predefined format of
VertexPositionTexture.VertexDeclaration. There are various different
predefined formats and its also possible to create custom user defined formats, for more information have a look
here. I realise I’m glossing over a lot of information, this is because the field of 3D graphics is huge even an API
such as XNA/MonoGame which tries to simplify things, there is still a vast array of different concepts and I don’t want to
get too bogged down with all the specifics.
vertexBuffer is assigned to the graphics device:
loads the vertex buffer into the graphics card ready to be draw later.
Next we move on to the
1 2 3 4 5 6 7 8 9 10
Update method is called every time the game decides that game logic needs to be processed. This includes the
management of game state, the processing of user input, and also the updating of simulation data or AI.
First of all we check the Escape key has been pressed so that the application can exit:
if Keyboard.GetState().IsKeyDown(Keys.Escape) then x.Exit().
Next we capture the amount of elapsed time since the last update so that we can calculate distance moved over time etc.
To make our view of the world less static we create a rotation around the z axis of the world so that we see the tetrahedron
from different angles. We multiply the rotation matrix by our initial
Matrix.CreateLookAt... that we used earlier on, and
assign it back to the
View property of the
basicEffect. I want to stress that the aim of this is not super optimal code
it’s merely to show the easiest possible method of achieving a result. In a future post we will be looking at some functional
scaffolding to allow us to apply functional thinking to this domain. Perhaps introducing a small Domain Specific Languagee to help.
Finally we have the
1 2 3 4 5 6 7 8 9 10
Draw override is called every time the game needs to draw a frame, we put all out rendering code in here.
The first step is to clear the screen to a nice blue colour:
To draw our tetrahedron we need to loop through the different techniques in out shader (In this instance our basicEffect only has 1), apply the technique, then draw out triangles. You might remember earlier to created a the
vertexBuffer and assigned it to the graphics device. All we have to do is tell MonoGame that we want to draw 4 triangles and they are in a
That’s it all done! Well almost, now lets backtrack slightly and look at how we build the vertices for that tetrahedron.
Building a tetrahedron
What is a tetrahedron? Well if you look on wikipedia
A tetrahedron is a polyhedron composed of four triangular faces, three of which meet at each vertex. It has six edges and four vertices. The tetrahedron is the only convex polyhedron that has four faces.
In the case of a tetrahedron the base is a triangle(any of the four faces can be considered the base), so a tetrahedron is also known as a “triangular pyramid”.
For any tetrahedron there exists a sphere (the circumsphere) such that the tetrahedron’s vertices lie on the sphere’s surface.
The tetrahedron is also the simplest of the five platonic solids. There are lots of interesting properties of these but I don’t really want to go into that here we just want to draw and texture one for now.
So how do we construct a tetrahedron?
There are various methods that can be used to construct a tetrahedron ranging from formula such as:
Cartesian coordinate based (±1, 0, -1/sqrt2) (0, ±1, 1/sqrt2)
V1=(2sqrt2 /3, 0, −1/3)
V2 =(− sqrt2 /3, sqrt6 /3, −1/3)
V3=(− sqrt2 /3,− sqrt6 /3,−1/3)
Yes I know I need to get latex maths expression working in my blog! Ill have to work on that.
I don’t know about you but I always feel uneasy unless I can clearly see exactly what’s been done, I also don’t want to turn this into a 3D geometry lesson because that’s not what I intend this post to be about.
Here’s what works for me anyway.
Calculate the radius of the circumsphere, this is the sphere in which all of the vertices of the tetrahedron sit, this is calculated by sqrt 3/8.
The angle between each vertex and its centre point is acos -1/3 or ~ 109.471 degrees.
- The first vertex is (0, sqrt (3/8) * length, 0)
- To get our second vertex we need to rotate the first vertex by acos -1/3 in the X axis
- To get the third vertex we rotate second vertex by 120 degrees in the Y axis
- For the last vertex we rotate the second vertex by -120 degrees in the Y axis
A picture can often be a worth a thousand words, I think this is one of those times, I will refer you to Friedrich A. Lohmüllers site for an excellent pictorial and description. The code for this process is below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
The last piece of the puzzle is the texture coordinates. There is some amazing software available to help model both texture and 3d geometry, projecting the vertices onto a 2d plane can be an art-form in itself. Luckily the tetrahedron is one of the simplest models, if you imagine the tetrahedron unfolded it would look like this from the top:
To map a texture to the tetrahedron we have to include a texture coordinate with every vertex. These coordinates are
in the code above. We use some ratios to select the correct coordinates within the texture. The texture coordinates are always
between 0 and 1. Here’s the location of the above points so you can see the locations clearly.
To make sure that the texture is in the right place we are going to use a type of fractal called the Sierpinski triangle. The Sierpinski triangle had exactly the same net, or unfolded shape as the texture we need to use. Each of the first iterations of the fractal is coloured separately as this will make it easy to see if the mapping is correct. Here is what the texture looks like:
This is how everything will look, I know its not incredibly impressive but its MonoGame in 3d, F#, and all running on a Mac, what more could you want! ;)
Its feels like we covered a lot of ground here but all we have is a spinning tetrahedron, its tricky to know what level of detail to go down to. I don’t want to teach anyone how to suck eggs, and I want to alienate people who are new to this area and want to learn, I hope I got the balance about right.
If you want to just get the code and have a look then here’s my GitHub repo
As usual I appreciate any comments and feedback.
Until next time!