It is generally easy to go from a vector graphic to a bitmap (the process of `rasterization’). Similarly, it is straightforward to go from a high-resolution bitmap to a lower-resolution one (the process of `downsampling’). The latter is common when you are preparing a bitmap for use on a Web page, intended to be viewed on a monitor. Careful downsampling can avoid the striping (`Moiré’ effect) sometimes seen in bitmap images. On the other hand it is generally much harder, and very unsatisfactory, to go in the other directions. It is possible to synthesize a bitmap graphic of `higher’ resolution by interpolation, but no new information is actually present. (This sleazy trick is occasionally disguised by the phrase `effective resolution’ rather than the true `optical resolution’.) This is what happens with the “Digital Zoom” on digital cameras. Converting a bitmap image into a vector image generally requires mathematically processing the bitmap to identify contours of constant intensity and/or or color (`quantization’) and then placing lines of appropriate thickness to delimit these regions. The resulting file is a genuine vector graphic but generally has a more restricted color and/or intensity range because of the mathematical processing. It is also often significantly larger than the original bitmap image. Source.