Abstract The interrelationship between variables involved in focusing schlieren systems is fairly well understood, however how changing the variables affects the resultant images is not. In addition, modified grids and arrangements, such as two dimensional, colour and retroreflective systems have never been directly compared to a standard system. The existing theory is developed from first principles to its current state. An apparatus was specifically designed to test grid and arrangement issues while keeping the system geometry, optical components and the test object identical. Source grid line spacing and clear line width to dark line width ratio were varied to investigate the limits of diffraction and banding and to find an optimum grid for this apparatus. Two dimensional, colour, retroreflective and a novel projected arrangement were then compared to this optimum case. In conclusion, the diffraction limit is accurately modelled by the mathematical equations. The banding limit is slightly less well modelled as additional factors seem to affect the final image. Inherent problems with the two dimensional and colour systems indicate that while they can be useful, they are not worth developing further though chromatism in the system meant that colour systems were not fully investigated. The retroreflective and projected systems have the most potential for large scale use and should be developed further.