Some food for thought... Just for the sake of discussion (and to make it easy), let's assume that on Mike's head texture, the UV map is split front/back and each take up 50% of the map. So on a 4000 pixel wide map, the front took up 2000 pixels of that and the back of the head took up 2000 pixels. Ok, so let's assume that the textures you create are indeed 4000 pixels wide (2000 front and 2000 back)... on an image like above, where the front of his head takes up less than half the width of a 640 pixel wide image (let's just say that it takes up 400 pixels, just to make it easy), then the software has to average 2000 pixels of texture into a 400 pixel space. Meaning that you can actually get 'less definition' in your image, because you throw out (or muddy up with averaging) 4 out of every 5 pixels of texture/bump. On the other hand, if you render a 1000 pixel wide image of half of his face (something like centered up, with maybe the edges of his eyes on the edges of the image), you'd get a 1-to-1 (roughly) texture->image ratio. In short, unless you're going to be rendering large images of extreme close-ups, 4000 is overkill. On the other hand, if you're creating textures that others will use, and you don't know how they intend to use them, it's always easier to scale down than up, so bigger is better unless there are filesize issues. Personally, I almost never use textures over about 3000 and I'd rather buy nicely detailed (and lower compression setting) 2000 sized textures than highly compressed 4000 sized textures when it comes to dimension vs. filesize vs. image quality.

Cinema4D Plugins (Home of Riptide, Riptide Pro, Undertow, Morph Mill, KyamaSlide and I/Ogre plugins) Poser products Freelance Modelling, Poser Rigging, UV-mapping work for hire.