Shaders, let's shed some light in this.

In 3D shaders are extremely important in giving polygons more information and to flesh out an object, rather than it being just a simple flat image.

When a polygon starts out, it has no values or anything overly distinguishable other than often a single colour. Shaders add the details, colouring and material to the polygons and without them they will be mostly lifeless.

The 3 main shaders used are:

Colourmaps
Specular maps
Bump/Normal maps.

Colourmaps are exactly as they are implied. They add colouring to the polygon surface, whether it is a hand painted texture or a photograph. The colourmap sets the overall colour of the object.

Specular Maps are another layer added to the colourmaps to bring out reflective highlights. The highlights are depicted by a black and white image, where the whitest values are the lightest/most reflective and the blackest values are the dullest/ least reflective.

Bump/Normal maps are a shader used to give depth to the colourmap and show details in the surface. A bump map works similar to speculars, in which the whites are higher surface peaks and black is a deeper surface. Normals are a more accurate version of bump maps and allow for greater detail, but rather than using black and white it uses Red, green and blue allowing more angles to be picked up. Generally normal maps are extracted from higher polygon meshes and baked onto a lower poly version giving the illusion of the lower poly model having more detail than it does.

If we were to make a golf ball, we would use a basic white texture for the colourmap, a patterned bump map of small circles repeated and then lightly highlighting the ball with a specular with a light amount of noise/dither to bring out the shine ( or lack of shine ) on the ball. To add even more detail, a secondary colourmap can be added with dirt and marks. The dirt and marks can then be softened/randomised by creating an alpha channel and painting white/black for where the ball should be dirtier/scuffed.

Talking Topology

Topology is a term used in 3D graphics to describe how a mesh is sculpted and the overall cleanliness of the polygons. Essentially it comes down to using polygons correctly and not wastefully, ensuring they are readable, ensuring they are optimised and ensuring there are enough for objects to bend.

By using good topology;

- The mesh is easier to read
- The mesh flows correctly to the shape of the object
- The mesh can use less expensive polygons
- The mesh is easier to unwrap for texturing
- More polygons in the right places will ensure natural movement and bending of polygons
- Less wasted resources
- Something boxy does not need as many triangles as something organic as the shapes are easier to define, so polygons can be spared.
- Models design with animation in mind will work correctly.
- Shapes are correct, contours flow together.

Here is an example from clevermusings.com of good vs bad topology. As you can see the polygons on the left model are poorly placed and do not flow very well, whereas the model on the right is clean and uses polygons correctly to fit the shape of the face without resorting to using more polygons to fill in the gaps.

Fixing what is broken.

Sometimes things do not go as planned or a model does not work as you'd hope, so re-doing it is all you can do to fix the problem.

I was not happy with the initial head, so I re-did it from scratch

Well within budget and matches what I wanted the character to look like. Happy days.

Short animation.

Well now, this is charming.

Uncharted 4 model

When looking for extremely high quality 3D character models, I often look at the works of Naughty Dog. As technology has improved, they have pushed realism in their models, especially in their Uncharted franchise.

                                For a real time model, this is amazing.

Polycount 1000 poly challenge

 On the polycount forum, there is a thread that allows creators to show their low-poly modelling skills. By sticking with a low poly count, the creator can give more room for processing power and texture space, which is essential in game development.

Some of the models on display are fantastic and are even viewable via Unity web plugin.

Polycount Forum

Check it out.

Character Modelling, Rigging and animation project.

For my current project, I will be creating modelling and rigging a character for animation.

Firstly the character is designed:

Front side and head profiles are created, with these a 3D mesh is extruded to create the shapes.

Using the 4 windows of Maya, I extrude a square primitive from the torso outwards, adjusting the vertex loops to fit the shape of the 2D profiles, while ensuring the minimum amount of triangles. Quads are preferred when creating models.

Using Mudbox, I create a higher polygon model from the low poly base model, this high poly model will be extracted as a normal map to be applied as a difference on the Low poly model, giving the illusion that a low poly model is displaying the same amount of detail as the high poly model.