We theoretically demonstrate that some doublets of NH 3 broadened by Ar and heavier atoms may be suitable for the first experimental demonstration of a so-far unstudied problem: The spectral effects of the speed dependence of line-mixing. By using realistic assumptions and spectroscopic data from previous studies, we show that neglecting this process leads to errors on the spectral shape of up to 2% of the peak absorption value. When multispectrum fits are made assuming speed-independent line couplings, the peak-to-dip residuals amplitudes reduce to about 0.5% and 1% for NH 3-Ar and-Xe, respectively. The magnitude of the effect is thus comparable to that of the speed dependence of the line broadening on isolated shapes, which has been demonstrated in many experimental studies. It should hence be detectable with high accuracy modern laboratory spectroscopic techniques. With this aim, guidelines and conditions paving the path for future experiments are given.