Generally stellar spectra, such as those obtained by SONG are analyzed assuming that the effects of the oscillation modes on the spectra are well approximated by a simple Doppler shift. As shown in Schou (2018) this is not a very good approximation, especially for Stars rotating much faster than the Sun. The result is that there is a high risk of systematic errors and that much of the useful information in the spectra is left unused.
Here will briefly review the results in Schou (2018) and discuss how they relate to the analysis of SONG data.