Influence of separating radii R₀ on marginalised chi-squares χ_norm² = (χ² − χ_min²)/Δχ_CL² versus model-recovered bar pattern speeds Ω_p. From left to right: a bar side-on case Au-18-89-50, a bar side-on case Au-28-85-80, and a bar end-on case Au-23-85-30 under the strong bar assumption (p_bar ≤ 0.50). Each fold line represents a complete set of modelling runs. The black fold lines denote the model results with default separating radii (R₀ = R_0,default), as previously displayed in Figs. 7 and 8, while the blue and red lines represent results with smaller (R₀ = 0.6R_0,default) and larger (R₀ = 1.3R_0,default) separating radii, respectively. The horizontal dashed lines represent 1σ (68%, χ_norm² = 1) and 2σ (95%, χ_norm² = 2) confidence levels. The vertical dotted lines denote the true bar pattern speeds, with their values shown by the green annotations.