Skip to main content
Commonmark migration
Source Link
  1. A "face" simply specifies one of the six images at a level in a cubemap (and its optional mip pyramid).

    A "face" simply specifies one of the six images at a level in a cubemap (and its optional mip pyramid).

    It is helpful to think of level L of your mipmapped cube array texture as consisting of a 2D array of images, with exactly six columns comprising the six faces of the logical cube, and the N rows corresponding to the N layers of the array (some APIs call these "slices"). This is because non-array Cube textures are already structured like 2D arrays with exactly 6 layers.

  2. This is correct with the exception that it would refer to the entire mip pyramid for that layer-face. Regarding the 2D array of faces and layers i just mentioned, OpenGL uses a single index for this 2d array, "unrolls" it, and calls this 1D index a "layer-face".

  3. Yes, but again, encompasses the mip pyramids for each layer-face image in that layer

  4. Depends on the context, could be either.

It is helpful to think of level L of your mipmapped cube array texture as consisting of a 2D array of images, with exactly six columns comprising the six faces of the logical cube, and the N rows corresponding to the N layers of the array (some APIs call these "slices"). This is because non-array Cube textures are already structured like 2D arrays with exactly 6 layers.

  1. This is correct with the exception that it would refer to the entire mip pyramid for that layer-face. Regarding the 2D array of faces and layers i just mentioned, OpenGL uses a single index for this 2d array, "unrolls" it, and calls this 1D index a "layer-face".

  2. Yes, but again, encompasses the mip pyramids for each layer-face image in that layer

  3. Depends on the context, could be either.

  1. A "face" simply specifies one of the six images at a level in a cubemap (and its optional mip pyramid).

It is helpful to think of level L of your mipmapped cube array texture as consisting of a 2D array of images, with exactly six columns comprising the six faces of the logical cube, and the N rows corresponding to the N layers of the array (some APIs call these "slices"). This is because non-array Cube textures are already structured like 2D arrays with exactly 6 layers.

  1. This is correct with the exception that it would refer to the entire mip pyramid for that layer-face. Regarding the 2D array of faces and layers i just mentioned, OpenGL uses a single index for this 2d array, "unrolls" it, and calls this 1D index a "layer-face".

  2. Yes, but again, encompasses the mip pyramids for each layer-face image in that layer

  3. Depends on the context, could be either.

  1. A "face" simply specifies one of the six images at a level in a cubemap (and its optional mip pyramid).

    It is helpful to think of level L of your mipmapped cube array texture as consisting of a 2D array of images, with exactly six columns comprising the six faces of the logical cube, and the N rows corresponding to the N layers of the array (some APIs call these "slices"). This is because non-array Cube textures are already structured like 2D arrays with exactly 6 layers.

  2. This is correct with the exception that it would refer to the entire mip pyramid for that layer-face. Regarding the 2D array of faces and layers i just mentioned, OpenGL uses a single index for this 2d array, "unrolls" it, and calls this 1D index a "layer-face".

  3. Yes, but again, encompasses the mip pyramids for each layer-face image in that layer

  4. Depends on the context, could be either.

Source Link

  1. A "face" simply specifies one of the six images at a level in a cubemap (and its optional mip pyramid).

It is helpful to think of level L of your mipmapped cube array texture as consisting of a 2D array of images, with exactly six columns comprising the six faces of the logical cube, and the N rows corresponding to the N layers of the array (some APIs call these "slices"). This is because non-array Cube textures are already structured like 2D arrays with exactly 6 layers.

  1. This is correct with the exception that it would refer to the entire mip pyramid for that layer-face. Regarding the 2D array of faces and layers i just mentioned, OpenGL uses a single index for this 2d array, "unrolls" it, and calls this 1D index a "layer-face".

  2. Yes, but again, encompasses the mip pyramids for each layer-face image in that layer

  3. Depends on the context, could be either.