Question

I am IT student and I have to make a project in VB6, I was thinking to make a 3D Software Renderer but I don’t really know where to start, I found a few tutorials but I want something that goes in depth with the maths and algorithms, I will like something that shows how to make 3D transformations, Camera, lights, shading …

It does not matter the programing language used, I just need some resources that shows me exactly how to make this.

So I just want to know where to find some resources, or you can show me some source code and tell me where to start from.

Or if any of you have a better idea for a VB6 project.

Thanks.

Answer 1

Many years ago I made a shaded triangle renderer that used library calls to draw the triangles. It’s a rather naive approach but you would be able to achieve the same result using VB6. I got all the maths & techniques from “Computer Graphics principles and practice” by Foley et al. Some parts are out of date now but I think you’d find it very helpful for this project and it can be bought 2nd hand at reasonable prices from Amazon for example.

One simple approach could be:

1. Read model file as triangles
2. Transform each triangle using matrices to account for camera position
3. Project triangle points onto 2D
4. Draw 2D triangle (probably using GDI)

This covers wireframe viewing. To extend this to hidden surface removal you need to work out which triangles are in front. Two possible ways:

1. Z-order sorting the triangles and drawing the ones furthest from the camera first. This is simple but inefficient if there are a lot of triangles and can give overlapping triangle effects when the order is not quite correct. You also have to decide how to sort the triangles – e..g by centroid, by extents…
2. Using a software depth buffer. This will give better results but is more work to implement. You will have to write your own triangle drawing code so cannot rely on GDI. See bresenham’s line algorithm and related algorithms for doing filled triangles for how to do this.

After this you’d also need some kind of shading based on lighting. The calculations are covered in Computer Graphics principles and practice. For simple shading you can stick with drawing triangles using gdi , but if you want to do gouraud or phong shading the colour values vary across a triangle. One way around this is to sub-divide the triangle into smaller triangles, but this is inefficient and won’t give very nice looking results. Better would be to draw the triangles yourself as required above for the software depth buffer.

A good extension would be to support primitives other than triangles. Basic approach would be to split primitives into triangles as you read them.

Good luck – could be an interesting project.