A new 3D rendering style for solid wire-frames

Now I'm improving one of my new 3D rendering engines by adding new features. It's one more set of improvements for my vector based drawing/rendering library named Maccala.

I'm doing all of this because I have great plans. In fact, I'm developing the functionalities I will need to create new graffiti text effects.

There is a new feature that is almost ready to use. I called it "clean solid wire-frame rendering". It helps to render 3D text that has just the essential lines in the thickness dimension.

It can seem to dummies it's just a trivial simplification, but it's not. It was very hard to develop because the shapes need to be drawn solid and opaque (it goes beyond the technique called "hidden lines removal"). My other solid wire-frame rendering algorithm already was very complicated and strongly based on heuristics. This "clean up" of edges invalidated most of tricks I have used before and needed a lot of other ones. It was almost a total redevelopment involving very hard work.

Below I show this new 3D rendering algorithm in practice. First you see a "normal" 3D wire-frame text and so a "cleaned" 3D wire-frames text created by using my new drawing style.

 Subscribe in a reader

Transforming 3D objects

Now Picture to People 3D libraries can make geometric transformations over some kinds of complex 3D objects.

After some more development effort it will be possible make transformations over the camera.

Above you can see an "a" letter after applying some geometric transformations.

Subscribe in a reader

Complex 3D objects

Now Picture to People 3D libraries can deal with complex 3D objects. By "complex" I mean objects composed by two or more simple objects, and they can even have holes.

An important point is the libraries can correctly represent the holes normals of these complex 3D shapes.

Above you can see a 3D "a" letter in wire-frame correctly drew with its hole and all its important normals.

Subscribe in a reader

Picture to People in 3D is evolving

Picture to People 3D libraries are already able to make a correct mapping of some kinds of 3D objects that can not be represented by a implicit or explicit equation (usually equations make calculations and representations simpler).

The more important is the normal vectors directions are calculated correctly. Sometimes a big problem with 3D models is that it's hard to know when you are inside or outside a complex object.

Above you can see a 3D object drew with its normals.

Subscribe in a reader

Picture to People in 3D world

I decided to take P2P project strongly to 3D world. I'm planning a very powerful 3D rendering engine for Picture to People.

Since until now P2P deals very few with the "Z-axis", I needed to start a new sub-project for the core libraries. Basically there is nothing ready for this, so I will need to make all the low level 3D libraries from scratch.

I'm projecting this engine very carefully, since I wanna make it very clean and extensible. Possibly I will base it in some ray-tracing algorithm variation, but I'm still considering other alternatives like radiosity. In fact, probably it will be a merge of several techniques.

For now the main practical goal is to make a 3D text renderer able to generate awesome results. Since I already started a draft of some low level 3D libraries, above you can see a 3D letter in wire-frame already made with Picture to People.

Subscribe in a reader