Carbon–proton dipole–dipole couplings between bonded atoms represent a popular probe of molecular dynamics in soft materials or biomolecules. Their site-resolved determination, for example, by using the popular DIPSHIFT experiment, can be challenged by spectral overlap with nonbonded carbon atoms. The problem can be solved by using very short cross-polarization (CP) contact times, however, the measured modulation curves then deviate strongly from the theoretically predicted shape, which is caused by the dependence of the CP efficiency on the orientation of the CH vector, leading to an anisotropic magnetization distribution even for isotropic samples. Herein, we present a detailed demonstration and explanation of this problem, as well as providing a solution. We combine DIPSHIFT experiments with the rotor-directed exchange of orientations (RODEO) method, and modifications of it, to redistribute the magnetization and obtain undistorted modulation curves. Our strategy is general in that it can also be applied to other types of experiments for heteronuclear dipole–dipole coupling determinations that rely on dipolar polarization transfer. It is demonstrated with perylene-bisimide-based organic semiconductor materials, as an example, in which measurements of dynamic order parameters reveal correlations of the molecular dynamics with the phase structure and functional properties.