In a previous work (Eff-Darwich & Pérez-Hernández 1997), it was presented a new inversion method based on the Regularized Least Squares technique. This new method, called OMD (Optimal Mesh Distribution) finds the optimal mesh and smoothing function by means of a deep analysis of the basis functions of the inversion problem. In the present work, the spatial resolution analysis developed for OMD to find the optimal mesh, has been adapted to a different inversion procedure, the SOLA (Substractive Optimal Local Averages) method (Pijpers & Thompson 1994). With the modification we have implemented in SOLA, the time and work necessary to obtain the solution have been minimised. Such modification consists on defining the width of the averaging kernels (critical point in SOLA) from the spatial resolution derived for OMD. The modified SOLA method has been successfuly applied to structure inversions. A comparison with other inversion techniques allows us to study the effects of the inversion procedure in the solution. We are mainly interested in the effects of the treatment of the surface uncertainties and the way the inversion technique deals with the solution in the core.