Glass

Glass render. Will be adding caustics soon.

Creating the Laboratory

I have used several Vyonyx textures to visualise a basic environment for the machine to dwell in:
















Adding the glass container led me to the idea that I can use caustics in this scene, putting mental ray to good use, and perhaps have liquids inside the flasks. I will most likely model these using surface meshes as they are background objects, and their shape lends itself to NURBS modelling.

Also noticable is the ominous shadow the robot casts, the wires being the main cause of this.

As you can see, my robot is still awaiting rigging...

http://www.vyonyx.com/index.php

Chest Piece - Update

After a suggestion by a friend I have made some of the pipes more angular. The torso now has a brushed steel texture applied to it, although I would ideally like to generate this procedurally.

Photorealism

This is where I hope to be by the end of this project:















If I can recreate this image perfectly, then one of my goals will be achieved.

Chest Piece

The chest piece is coming along nicely - the reflections from the pipes add an eerie high tech alien feel to the machine. I'm still not happy with the polygon flow in the centre - there seem to be unwanted ridges even with high levels of subdivision.

Subdivision Spree!

I went on a subdivision spree to improve the look of the robot. Here's the lower face:

Simulation Test

I did a quick test in XSI to see if I could make a brick wall collapse:



Observations:

Zero Gravity looked cool but was unrealistic. The robot is not that strong. Bricks cannot spin in mid-air.
Having nothing to 'push against' in the first test, the wall exploded outwards and upwards as well as the intended direction, forwards.

Adding passive rigid body bricks to the sides of the wall give it increased stability and believability, and prevent the wall from shaking itself apart before the 'fist' has collided.

The wall tends to 'dip' at start of simulation - perhaps I should reduce amount of active rigid bodies? 'Mortar' in between bricks may fix this problem. Another potential reason may be the amount of subdivisions on the 'floor' is not great enough to hold up the wall, but this is unlikely.

Realistically there will only be certain sections of the wall which will be destructible - the rest will be rigid - possibly round holes for the fist to smash through. The hole needs to be big enough for us to see the robot from the other side.

Another problem will be preventing the entire centre of the wall from falling down - perhaps steel girders or struts could spererate wall partitions.

I will also have to render this scene without breaks, or pre-cache the simulation, to prevent any changes to the animation occurring.

Ambient Occlusion Test

Overall I am happy with how this has come out - the cogs and gears are emphasised nicely. There is still a lot of 'graininess' with 100 samples, so I will increase this for the final render.

Knees

An important design change to the knees gives the robot realistic flexibility. I will probably change the arms as well.

3D Robot

Newest render of the 3D Model. Added cogs to the chest to link it to the head.

Initial Concept Idea

This is the basic idea for my project - an animation focused around a robot made from spare parts - Frankenstein's robot...