Brave New World

A demonstration of POWERVR MBX™'s advanced skinning effects. Naked denizens of the Brave New World balance on the edge of uncertainty. The one thing we can know for sure is that with per-pixel DOT3 Lighting and skinning using 24 bones per character they look great.

View Video Demo (BraveNewWorld.m1v, 5 MB)

Back to Top

Butterflies

This demonstration shows a high number of butterflies in a dynamic flock illustrating the alpha blending capabilities and performance of POWERVR MBX™.

There are more than 30 different butterfly textures all compressed down to only 2 bits per pixel using POWERVR texture compression.

View Video Demo (butterflies.m1v, 5.8 MB)

Back to Top

Racer

POWERVR Racer is a high-speed, white-knuckle car race demo with contestants battling through an entire city. Racers tear downtown into the industrial quarter and the mountains beyond, then back through the sleepy suburbs to the commercial district.

Although this is a 'canned' demo with no interaction, significant game-like code is executed, with banking, braking and suspension being simulated in real-time by the CPU, and executed entirely with fixed-point arithmetic. The city database is modelled with over 20,000 triangles, and the cars are modelled with multiple levels of detail, with 1,500 triangles in the most detailed version.

Accurate reflection-mapping and multi-texturing is used on the cars, PVR-TC 4-bit texture compression is used throughout.

View Video Demo (car_racer.mpg, 13.6 MB)

Back to Top

Chameleon Man

The Chameleon Man demo shows a matrix skinned character in combination with DOT3 per pixel lighting.

Matrix skinning is the act of animating a vertex over time given a set (palette) of matrices and a known set of blend weights assigned to those matrices. The Chameleon Man model has 19 bones and an animation cycle of 16 frames.

For each frame the matrix palette is recomputed based on time. For example, to render the model at time point 5.25 the application would linearly blend between the matrices stored for frame 5 and 6 using weights of 0.75 and 0.25 respectively.

Per vertex, up to 3 matrices, from the palette, along with 3 weights are used by the vertex program to update the vertex position to obtain the current animation frame position.

To implement DOT3 per pixel lighting (bump mapping) the application supplies a normal, binormal and tangent per vertex. These tangent space vectors are transformed by the vertex program into world space. The light vector is then transformed into this local skinned tangent space. The tangent space light vector is then used to do the DOT3 lighting calculation which is finally combined with the base texture using multi-texturing.

A standard vertex lit mode is supported to show the stunning visual difference made by per pixel DOT3 bump mapping.

View Video Demo (chameleon.mpg, 1.4 MB)

Back to Top

Fixed Point

This suite of three demonstrations relies only on fixed-point arithmetic.

Henge, the stone circle demo, uses fixed point application lighting, passing only RGBA vertices to the OpenGL pipeline.

The Hurricanes demo features seven WWII fighter planes animated entirely using fixed point arithmetic with PVRTC compression and dynamic lighting. In addition, both the fighter planes and hierarchical terrain model are actively culled to the view frustum to manage scene complexity.

The POWERVR Torso (RefMap) demo renders a geometrically complex model with accurate reflection mapping. All reflection mapping calculations are performed in fixed point with 2 layer multi-texturing and about 200K polygons per second.

View Henge Video Demo (fixedpoint_henge.mpg, 2.6 MB)

View Hurricanes Video Demo (fixedpoint_hurricanes.mpg, 3.4 MB)

View Torso Video Demo (fixedpoint_refmap.mpg, 2 MB)

Back to Top

Fur

Demonstrates a method for rendering fur. Multiple translucent "shells" of the original model are rendered, each slightly larger than the previous.

The texture applied to these layers contains a "dot" in the alpha channel to show where a hair passes through the "shell."

View Video Demo (fur.mpg, 1.7 MB)

Back to Top

Mars Base

The exploration of space is underway with POWERVR's first mission to mars. Pushing close to 20,000 polygons per frame the complex spaceship model also features a multi-textured reflection effect.

Video Demo Coming Soon

Back to Top

Mouse

Cell-shading is the "art" of rendering your objects to look like cartoons. This demo shows a dancing mouse with a hard black outline and banded shading to achieve the desired cartoon-look. Both effects are set-up using an optimised vertex shader program or, in the case the hardware platform does not have a VGP unit, using optimised software routines.

View Video Demo (mouse.mpg, 1.2 MB)

Back to Top

OpenVG

The POWERVR MBX™ and POWERVR SGX™ families provide full accelerated support for OpenVGT, the royalty-free, cross-platform API that provides a low-level hardware acceleration interface for vector graphics libraries such as Flash and SVG.

OpenVG provides portable acceleration of high-quality vector graphics enabling compelling graphical user interfaces, navigation, flash-like animation, SVG Clip-art and 2D gaming on small screen devices - with POWERVR hardware acceleration enabling fluidly interactive performance at very low power levels.

View Video Demo (openvg.mpeg, 19.8 MB)

Back to Top

Particles

This demo shows a physics based particle effect displaying a total of 900 particles (600 real + 300 reflected).

View Video Demo (particles.mpg, 2.6 MB)

Back to Top

Shadow

This demo illustrates various shadow techniques.

The static blob method draws a simple static transparent blob under the character - this is the most basic shadow technique used in games.

The dynamic blob method draws a transparent blob under the character and the blob scales and moves based on the bounding box of the character, the bounding box is update each frame.

The projected geometry method draws the character "projected" onto the floor plane, only pure black shadows are possible due to multiple polygons being projected into the same location (would give intensity differences in the shadow if transparency would be used).

The final method is projected texture where the projected shadow texture is updated dynamically each frame (generated by rendering the scene from the point of view of the light source). This method, unlike the previous methods, also works correctly for shadows cast on non-planar objects.

View Video Demo (shadow.mpg, 3 MB)

Back to Top

Show Room

The Imagination Technologies people carrier transports over 14,000 vertices and illustrates POWERVR MBX™ anti-aliasing and multi-texturing capabilities.

View Video Demo (showroom.m1v, 2.8 MB)

Back to Top