Currently we just render the SVG at a fixed resolution (128x128 or 256x256) and then use gl.generateMipmap(gl.TEXTURE_2D) to generate mipmaps. This does not utilize vector information though leading to visible artifacts at the edges when repeating the texture. Perhaps this could be improved by generating the different levels of detail ourselves. The SVG file should be loaded just once and then rendered at different resolutions.
Currently we just render the SVG at a fixed resolution (128x128 or 256x256) and then use
gl.generateMipmap(gl.TEXTURE_2D)
to generate mipmaps. This does not utilize vector information though leading to visible artifacts at the edges when repeating the texture. Perhaps this could be improved by generating the different levels of detail ourselves. The SVG file should be loaded just once and then rendered at different resolutions.