Circuit and method for trilinear filtering using texels from only one level of detail

Abstract

A circuit and process perform trilinear filtering using four texels (called “nearest texels”) that are nearest to a to-be-displayed pixel, and also using twelve additional texels (called “surrounding texels”) that surround the nearest texels. The nearest texels and the surrounding texels (together called “fine texels”) are all from only one level of detail L, while a filtered texel being generated is at another level of detail L+p, wherein p is a fractional level of detail. The filtered texel is used in rendering the to-be-displayed pixel, and can be identical to the texel obtained by trilinear filtering in the prior art. The circuit and process use fine texels to regenerate a quad of coarse texels that are used with a quad of the nearest texels to perform trilinear filtering. Alternatively, the circuit and process generate coefficients from the S and T coordinate fractions, and multiply the coefficients with the nearest texels and with summed texels (obtained by adding three surrounding texels and optionally the nearest texel), and add the products to obtain the filtered texel. In one implementation, a coarse texel regenerated from a number of fine texels is compared with another coarse texel pre-existing in a mipmap at the level of detail L+1, and in case of a match additional coarse texels are regenerated for the trilinear filtering in a single cycle. In case of no match, trilinear filtering is performed in two cycles using pre-existing coarse texels of the L+1 mipmap.