The other day I wanted to get some good reflections on an object but didn’t want to place it in another scene. (I should have, it would have been faster. But at least I learned something new this way…) So I set up a scene with 6 cameras and rendered a cubemap to use as an environment texture for the new object. Turns out it’s not that simple as modo doesn’t quite like cubemaps on the environment. So I had to come up with a new way and created a spherical map, which worked out great…

Cubemap: set up 6 cameras and render an image with each one; assemble in photoshop…
Spherical Map: make a reflective sphere with a good uv layout and bake to texture…

remind me to go into more detail and post some screens…

Based on a bit of research, watching the Unite ’07 Performance Optimization talk and talking on the unity forums:

Each material on each mesh causes one draw-call per frame, per light that affects it and per camera that sees it.

If I have 1 mesh with 20 seperate textures, only ambient light, one camera, this will also result in 20 draw-calls. If I would combine the 20 textures into one atlas the scene would now only need a single draw-call.

If I have 20 meshes with one texture each, only ambient light, one camera, this will result in 20 draw-calls. If I would combine all the textures into one big atlas, the scene would still require 20 draw-calls, even though all the objects would now use the same texture.

The only way to optimize here would be to combine meshes. But you can only do that if your meshes aren’t to move around independently from each other.

PVRTC is a block-based texture compression, that means that it works with groups of pixels. Problems mostly show up in high contrast areas where too many different extremes meet.

“It uses a representation that is based on the blending of two ‘low frequency’ signals using a high frequency but low precision modulation signal.” So when compressing, it takes your image, creates two small blurry versions and a modulation image. When putting it back together it calculates the value of each pixel like this: It scales the 2 small images back up and samples the 2 colors from the right location, and then decides which shade in between the 2 values to take based on the modulation image.

There’s an example on page 4 here, just look at Figure 3 and Figure 2: http://web.onetel.net.uk/~simonnihal/assorted3d/fenney03texcomp.pdf
(This is the doc I quoted above)

So the solution is very simple, isn’t it? Just try not to use more than 3 different colors in each 4×4 pixel block… ;]

• A thread on the unity forum about this topic

Camera Position matters a great deal when baking Indirect Illumination! Look at the example below. – This is part of the ceiling of a room where the camera was on one end when baking. Notice how the quality deteriorates towards the right:

Once I moved the camera below the room, moved it a bit away from the scene and pointed it up, the result improved greatly. (Of course I had to adjust the Indirect Illumination settings a bit to get good results with the camera further away from the scene)

And while I’m at it:

A short summary of what the Irradiance Caching Options do for future reference (info from the modo user guide):

The difference between Indirect Illumination with or without Irradiance Caching is that “Irradiance Caching relies on much higher quality samples distributed sparsely across the project whereas Indirect Illumination without Irradiance Caching uses lower quality samples at every pixel.”

• Irradiance Rays: Number of rays fired
• Supersampling: After all rays have been fired, this fires 25% additional rays into difficult areas
• Irradiance Rate: Minimum distance between screen-pixels on the surface where irradiance is computed
• Irradiance Ratio: Ratio between minimum spacing (Irradiance Rate) and maximum spacing between computed values
• Interpolation Values: How many nearby values are required to interpolate? if not enough values are found, it forces the computation of a new value