Anti aliasing / anti aliasing Technology

Root cause of aliasing , It is impossible to correctly express continuous geometry with discrete pixels , The pixels at the edge are not smooth .

But when there are more pixels, the higher the resolution , The less obvious the aliasing phenomenon is . So the first way to anti aliasing is to use a higher resolution in rendering calculation , This method is called oversampling antialiasing (Super Sample Anti-aliasing, SSAA), But this method obviously has a high performance overhead .

As shown on the left , It is usually used to judge whether a point belongs to a triangular patch , It depends on whether the pixel center point is covered , This central point is also called the sampling point . Increase the number of sampling points , The color is averaged according to the number of covered sampling points , As shown on the right , This is it. MSAA Method .

MSAA To use continuous geometric information , In delayed rendering GBuffer There is no continuous geometric information , So usually MSAA Not applicable to delayed rendering

FXAA The image processing algorithm is used , The anti aliasing effect is achieved by detecting the image edge and smoothing it .

TAA Principle and MSAA Almost the same , Each pixel has multiple sampling points . But different from MSAA The way ,TAA (Temporal Anti-Aliasing)  Integrate the data of historical frames to achieve anti aliasing , This will spread the multiple samples of each pixel into multiple frames , The relative cost is much smaller .