This paper is an exploration of the behaviour of high-resolution transmission electron microscope (HRTEM) images at up to 1 Å resolution. The ultimate limits to HRTEM (structure) resolution and the manner in which strong scattering may lead to weak diffraction in heavy fcc metals are discussed. A resolution of 1·0 Å is somewhat better than the ultimate resolution presently achievable in a 400-kV electron microscope. In heavy metals, such as platinum, it is found that the lattice fringe contrast is very low in the  projection, but that fringe contrast may be improved by imaging in the  projection. For atomic resolution imaging of the heavy metals in the  projection a resolution of 1·2 Å is required. For the study of oxygen position in high-temperature superconducting (HTS) oxides a resolution of between 1·2 and 1·4 Å is required. At better than 1·2 Å resolution the thick crystal images in HTS oxides remain simple and are easily interpreted. At such resolution all atomic columns are separated for the HTS  projection and the dynamical diffraction effects improve the contrast of oxygen atoms relative to the metal atoms.