First, some background on why this occurs.
With out mipmaps, as you shrink your texture down, the tiny fringe of half-coloured texels supplying you with a pleasant anti-aliased border between two colored areas shrinks to virtually nothing. The possibility that any specific pixel’s texture pattern lands proper on that razor edge goes towards zero, and so most pixels find yourself trying up their texture from one facet or the opposite, 100% the color on the left or 100% the color on the appropriate, leading to that harsh stairstep transition with no mixing.
Auto-generated mipmaps assist with this by pre-generating smaller variations of the asset which can be pre-blended for quick lookups (and higher cache coherency which is nice for efficiency). However once they solely have a raster picture to go off of, they must make guesses concerning the true steady color area they’re purported to be approximating. Normally they will use a easy field filter or some barely extra subtle methods of averaging the colors already baked into your raster. That averaging tends to smear out your borders greater than a real supersample of the supply vector would, contributing to the slight fuzziness you see.
You’ll be able to enhance this considerably by making your individual mipmaps instantly out of your vector supply. Your vector modifying software program is ready to apply supersampling to the bottom fact steady picture to get a really high-quality downsample, which something working from simply the raster can solely approximate.
So as an alternative of exporting only a single high-res sprite, export a complete sequence, every half the dimensions of the earlier. You’ll be able to mix these pictures right into a mipmap chain in DDS texture format — there are plugins for Photoshop and different picture instruments to assist with this, or you should utilize DirectX utilities to do it manually.
Importing this into Unity, it’ll use your mipmaps as an alternative of guessing at its personal with recursive filters. Then anytime your sprite is displayed at precisely one of many resolutions you exported, and aligned with the pixel grid of the display, you will get crisp edges similar to your direct vector output. So for very best quality, attempt to export your sprite at a decision such that the standard sizes it seems in your recreation every carefully match the decision of one of many supply pictures within the mip chain. You’ll be able to even swap between totally different variations of the sprites for closeup and much photographs, if the sizes you want do not align properly with an influence of two relationship.
However in-between these sizes you may nonetheless run into hassle. By default, utilizing trilinear filtering, the GPU will common two close by mip ranges to estimate the appropriate color for every pixel at an in-between dimension. This may once more introduce somewhat fuzziness, each from the averaging, and since a type of ranges is just too low-res/blurred for the dimensions you are drawing at. You’ll be able to power the filtering to bilinear / select simply the closest mipmap to pattern from, which can introduce somewhat upscaling fuzz when scaled up and somewhat aliasing when scaled down from the closest native dimension you saved — and a visual pop throughout scaling because it switches sources discretely. Or you may apply a mipmapping bias to shift the lookup somewhat extra strongly towards the extra detailed mip, buying and selling somewhat additional aliasing for crispness, whereas protecting the transitions clean and pop-free.
