Image Map
Layer - Select a layer to use as an image map. This image will be mapped onto a sphere of infinite size.
How Layers are used in Horizon
When a layer is selected using the "Layer" popup menu, the referenced layer is used instead of the built-in gradient. When a flat 2D image is mapped onto a sphere shape, distortion will naturally occur, unless the image has been specially created or adapted for use with a 3D sphere.
Spherical Maps
A spherical map has been pre-distorted so that it will look correct when mapped onto a 3D sphere. There are two main benefits to using special spherical maps: they are usually made to have no seams at any point along the sphere, even at the poles; and sphere maps are built to be wrapped spherically and will not look warped or stretched.
An example of a spherical map
Finding Spherical Maps
You will find a number of vendors around the internet that sell spherical maps or equirectilinear maps. One good source of high-quality maps is Dosch Design at www.doschdesign.com.
Spherical maps are used for many different purposes and come in different types.
The following types of maps will work well in Horizon:
Non-spherical Maps
You can also use regular photos, stitched panoramas and matte paintings (or movies) as Image Maps. Just set the Coverage H and V to desired values, and set the Straighten control to 100 in the Orientation and Distort group.
H Coverage- Describes in degrees (0-360) how much the image map covers the view horizontally. A value of 360 means that the left and right sides of the image will meet on the far side of the sphere, and will create more horizontal distortion. For an animation using a non-spherical map where the camera does not move a great deal, you may achieve better results by limiting the horizontal coverage to create less stretching.
V Coverage - Describes in degrees (0-180) how much the image map covers the view vertically. A value of 180 means that the top and bottom of the image will touch the poles, creating more vertical distortion in a non-spherical map.
Sampling:
Your 2D image is mapped to a 3D surface. However, there is not a one-to-one correspondence between each pixel and each coordinate on the sphere, because the image is stretched to fit the spherical surface. Sampling compensates for this lack of perfect correspondence and interpolates the right pixel value at each coordinate.
There are two sampling methods available in Horizon.
Nearest Neighbor sampling is fast, because it picks the closest pixel value to the coordinate, but can result in a pixelated effect. The default setting of Bi-Linear sampling averages pixels around the coordinate. Bi-Linear is slower but produces better quality.