Target independent rasterization

• US 9,183,651 B2
• Filed: 10/06/2010
• Issued: 11/10/2015
• Est. Priority Date: 10/06/2010
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Hinitz, Shai, Patel, Amar, Boyd, Charles N., Pelton, Blake D.
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): PRINGLE-PARKER, JASON A
• Application Number:
  US12/899,125
• Publication Number:
  US 20120086715A1
• Time in Patent Office: 1,861 Days
• Field of Search: 345/613, 345/522
• US Class Current: 1/1
• CPC Class Codes: G06T 11/40 Filling a planar surface by adding surface attributes, e.g. colour or texture

Abstract

Methods, systems, and computer-storage media for target independent rasterization of an image. The target is the memory allocated for a rendered image within a graphics pipeline. Embodiments of the present invention allow the rasterization process'"'"'s sampling rate to be specified independently from the memory allocated for the rendered image. Embodiments of the present invention also allow the rasterization process to be executed at a rate that does not correspond to the memory allocated for the rendered target.

Claims

1. A computer-storage device having computer-executable instructions embodied thereon for performing a method of rendering antialiased two-dimensional graphics, the method comprising:

converting scene data into an input coverage mask by generating a set of geometric shapes tessellated from objects in the scene data;

receiving, by an application program interface (API), a first instruction that defines a render-target storage value in a graphics pipeline, wherein the render-target storage value is an amount of memory allocated to a rendered image;

receiving, by the API, a second instruction that defines the sampling rate for a rasterizer in the graphics pipeline, wherein the sampling rate is defined by the second instruction independently from the render-target storage value and the render-target storage value is defined by the first instruction independently from the sampling rate, and wherein the sampling rate defines a number of sub-pixels that the rasterizer assigns colors;

receiving a third instruction to configure a pixel shader to process the input coverage mask through the graphics pipeline based on the amount of memory allocated and the number of sub-pixels that the rasterizer assigns colors, wherein the pixel shader and an alpha blender are configured to count a number of bit sets in the input coverage mask, normalize the results to a real number, and add the real number to current contents of a render target;

wherein at least the first and second instructions comprise configuration instructions for one or more components in the graphics pipeline; and

communicating the input coverage mask to the graphics pipeline.

2. The device of claim 1, wherein the method further comprises receiving rendered scene data from the graphics pipeline.

3. The device of claim 1, wherein the method further comprises:

setting, by a first function of the API, the render-target storage value as the amount of memory allocated to the rendered image based on the first instruction that is received from a computer program for which the image is to be rendered; and

setting, by a second function of the API, the sampling rate for the rasterizer as the number of sub-pixels that the rasterizer assigns colors based on the second instruction that is received from the computer program.

4. The device of claim 1, wherein the render-target storage value is defined to be one times a screen size, and wherein the sampling rate is set to a value greater than 1 sub-pixel per pixel.

5. The device of claim 1, wherein the pixel shader and an alpha blender are configured to count a number of bit sets in the input coverage mask, normalize results to a real number, and add the real number to current contents of a render target when there is no overlap between the set of geometric shapes.

6. The device of claim 1, wherein, when there is overlap between the set of geometric shapes, the pixel shader and an alpha blender are configured to combine the input coverage mask with current contents of a render target using a logical process to determine which colors are added or subtracted from the render target.

7. The device of claim 6, wherein the logical process is one of a bitmap exclusive-or (XOR) operation and a bitmap OR operation.

8. The device of claim 1, wherein the render-target storage value is equivalent to one pixel per pixel in a rendered image and the sampling rate is set to 4 sub-pixels per pixel.

9. The device of claim 1, configuring, by the API, functions in at least one GPU driver based on the first and second instructions.

10. A method of instructing a graphic processing unit to perform target independent rasterization, the method comprising:

receiving a first instruction that defines a render-target storage value in a graphics pipeline running in a graphical processing unit, wherein the render-target storage value is defined by the first instruction independently from a sampling rate;

receiving a second instruction that defines the sampling rate for a rasterizer in the graphics pipeline in the graphical processing unit, wherein the sampling rate is defined independently from the render-target storage value and the sampling rate defines a number of sub-pixels that the rasterizer assigns colors in the rendering of the image; and

configuring a pixel shader to process an input coverage mask through the graphics pipeline based on the amount of memory allocated and the number of sub-pixels that the rasterizer assigns colors, wherein the pixel shader and an alpha blender are configured to count a number of bit sets in the input coverage mask, normalize results to a real number, and add the real number to current contents of a render target;

wherein at least the first and second instructions comprise configuration instructions for one or more components in the graphics pipeline.

11. The method of claim 10, wherein the the pixel shader and the alpha blender are configured to combine the input coverage mask with the current contents of the render target using a logical process to determine which colors are added or subtracted from the render target.

12. The method of claim 11, wherein the logical process is one of a bitmap exclusive-or (XOR) operation and a bitmap OR operation.

13. The method of claim 10, wherein the pixel shader and the alpha blender are configured to count a number of bit sets in the input coverage mask, normalize results to a real number, and add the real number to current contents of a render target when there is no overlap between a set of geometric shapes tessellated from objects in scene data that was converted into the input coverage mask.

14. The method of claim 10, wherein the sampling rate is greater than the render-target storage value.

15. The method of claim 10, wherein the render-target storage value is equivalent to one pixel per pixel in a rendered image and the sampling rate is set to 4 sub-pixels per pixel.

16. The method of claim 10, wherein the image rendered by the graphics pipeline is a two-dimensional image.

17. A system comprising one or more processors configured to perform a method of instructing a graphic processing unit to perform target independent rasterization, the method comprising:

receiving a first instruction that defines a render-target storage value in a graphics pipeline running in a graphical processing unit, wherein the render-target storage value is defined by the first instruction independently from a sampling rate;

receiving a second instruction that defines the sampling rate for a rasterizer in the graphics pipeline in the graphical processing unit, wherein the sampling rate is defined independently from the render-target storage value and the sampling rate defines a number of sub-pixels that the rasterizer assigns colors in the rendering of the image; and

configuring a pixel shader to process an input coverage mask through the graphics pipeline based on the amount of memory allocated and the number of sub-pixels that the rasterizer assigns colors, wherein the pixel shader and an alpha blender are configured to count a number of bit sets in the input coverage mask, normalize results to a real number, and add the real number to current contents of a render target;

wherein at least the first and second instructions comprise configuration instructions for one or more components in the graphics pipeline.

18. The system of claim 17, wherein the pixel shader and the alpha blender are configured to combine the input coverage mask with the current contents of the render target using a logical process to determine which colors are added or subtracted from the render target.

19. The system of claim 18, wherein the logical process is one of a bitmap exclusive-or (XOR) operation and a bitmap OR operation.

20. The system of claim 17, wherein the pixel shader and the alpha blender are configured to count a number of bit sets in the input coverage mask, normalize results to a real number, and add the real number to current contents of a render target when there is no overlap between a set of geometric shapes tessellated from objects in scene data that was converted into the input coverage mask.