Frequency perturbations scaled to the Sun, plotted as a function of the simulation frequencies. Squares are for cases 10, 12, and 13, diamonds are for case 11, and crosses are for case 15. The solid lines connect the result of cases 10, 12, and 13. The top curve shows the RGM cases; the bottom curve shows the GGM results. The results for the first zero crossing below 10 Mm were used (as opposed to a shallower crossing) in order to ensure that the physics is well represented by the classic equations. The effects of using different crossings are discussed in Appendix A. For the sake of comparison, the upper dashed black line shows the residuals (Sun minus model) from the RGM patched model of Houdek et al. (2017, model “B”) and the lower dashed black line shows the residuals from the GGM patched model of Ball et al. (2016). Regarding the latter, we note that we were unable to reproduce the frequencies from the model (obtained from Ball, priv. comm.) exactly, due to what appears to be numerical problems in the model. However, the differences were small enough so as to not affect the overall conclusion. The red dashed lines show the results from Fig. 5 of Jørgensen et al. (2018), while the blue lines show the RGM numbers shown in blue in Fig. 6 of Jørgensen & Weiss (2019). The corresponding GGM numbers are courtesy of Jørgensen (priv. comm.). The numbers from Sonoi et al. (2017) are not shown as they have deviations at low frequencies, which are indicative of an overall structural error. The vertical dotted line indicates roughly where leaks from other k_h start to appear within the fitting interval, as is discussed in the main text.