Michel Rohner

Author of New-Based Anti-Aliasing For CGI & Anti-Aliasing with MSAA vs ABAA

My Books

New Area-Based Anti-Aliasing For CGI & Anti-Aliasing with MSAA vs ABAA

  • Better Image Quality, with less Subpixel samples when compared to MSAA.
  • Proposed implementations can be scaled to 4 and 8 Subpixels, and even 16 and 32 subpixels.
  • Simpler implementation and lower system cost.
    Michel expects that this new approach will be widely accepted and result in improved quality and
    lower cost for future 3D graphics adapters and 3D RT CGI systems.

New Area-Based Anti-Aliasing For CGI

Anti-Aliasing with MSAA vs ABAA

Subpixel Rendering and MSAA

There are some brief descriptions of Anti-Aliasing at the following two site links:

What Is Anti-Aliasing and Which Type Should You Use?

Subpixel Rendering and ABAA

The author has written 2 books about anti-aliasing:

The author has written 2 books about anti-aliasing:

New Area-Based Anti-Aliasing for CGI (extended  version of book)

Early Real-Time CGI Systems

Links to Early Real-Time Computer-Generated Imagery (CGI) Systems:

Anti-Aliasing was a requirement for early CGI Systems.

Link Flight Simulation Demo, Video, DIG 2 Demo, 1984:

Evans & Sutherland ‘The Tactical Edge’

Part 1

Part 2

Comparison of ABAA8 vs MSAA8 with Tri-Fans

Thin triangles processed with 8 Subpixels using ABAA and MSAA are compared. Reference Figure 2.

These examples have been generated by a C-program. In these four examples, arrays of 8 thin triangles are displayed within 8×8 Pixel Spans. Each triangle is defined by 3 edges within an 8×8 Pixel Span.

It is not practical to show images with gray shades. When printed, the images can be easier to evaluate at first. But, because of the limitation of the printing process, the results could have been inconclusive. By showing the covered Subpixels or the covered Subpixel with character counts per Pixel instead gray shade, it is easier to evaluate and compare the ABAA vs MSAA approaches.

The algorithms used for Subpixel processing can be evaluated by using thin triangles, of size approximately 8 Pixels long and 1 Pixel wide at the base. The test uses groups of 8 such triangles organize into fans spreading over an angle of Pi/2 (90 degrees).

Test Cases Consisting of 8 Narrow Triangle Fans

For the comparison of ABAA with MSAA, the test cases consist of 8 thin triangles organized as a fan displayed across a Span of size 8×8 Pixels. In the fan, there are four triangles (a, b, c & d) with vertical edges eight SL long (VE) and four triangles (e, f, g & h) with horizontal edges eight Pixel long (HE). The triangle bases are roughly 1 Pixel wide.

Since the triangle tops are 0 Pixel wide and the bottoms are 1 Pixel wide, the number of Subpixel increments per Pixel from top to bottom of triangle is 8 steps.  The increment for each step should be 1/8 Pixel each.

Results for ABAA

For ABAA, in most cases, the increments are:
0, 1, 2, 3, 4, 5, 6, 7, or
1, 2, 3, 4, 5, 6, 7, 8

Results for MSAA

Some Good: For MSAA, the increments are similar to ABAA when the triangles edges are near Vertical (face a) or near Horizontal (face h). This is because the Subpixels are positioned according to the 8 Queen algorithm.

Some Bad: For other orientations, the increments are not constants with many double increments. There is also ‘hesitation’, when there are some steps with decrements.