The goal of the paper is to propose modifications to tested flat cascades that will suppress partial distortions of acquired interferograms in the region of blade leading edges. This improvement will allow more accurate determination of actual flow incidence angles of tested blade cascades, in particular for transonic or supersonic inlet flows. Application of physical probes for such tasks is always in question in transonic or supersonic flows. The paper is composed of three main sections: (a) introduction of the test facility, (b) presentation of the problem with examples, and (c) description of the experimental work. Recommendations for future flat cascade investigations is presented in the paper. The first section is devoted to the introduction and description of the High-Speed Laboratory of the Institute of Thermomechanis of the Czech Academy of Sciences. Attention is paid to the unique large-scale interferometer which is one of the principal research instruments here and which is routinely used for investigations of transonic compressor and turbine cascades. The instrument capability is illustrated by a series of images showing evolution of a sonic line in a transonic cascade as a function of the increasing inlet Mach number. The reasoning for the proposed work is presented in the middle section. The major impetus for the work was to understand the observed discrepancies between schlieren and interferometer images while testing highly-loaded transonic compressor cascades. In particular, the main concern is the relatively wide region of increasing pressure in the shock vicinity recorded on interferograms versus sharp shock wave image visible on schlieren images. It was suggested that these discrepancies are caused by deformation of the shock-wave surface by the growth of secondary flow due to the tunnel endwall effects. It should be stressed here that the intention was not to investigate the pattern or the nature of the secondary flows rather. An idea behind this approach is to move the secondary flows out of the region of interferometer imaging. Finally, in the last section the results of the experiments carried out during the course of this work are presented. The experiments were designed to improve understanding of the origins of interferogram distortions. Further intention was to eliminate or at least lessen the level of interferogram distortions due to the combined effects of the boundary layer interaction and the corner-vortex flow. Wedges of a constant vertex angle of 15 deg of various plane shapes were inserted subsequently in supersonic flow of (Mach number 2) and interferograms of the resulting flow pattern were acquired. It was observed that decreasing the wedge span led to clearing the interferograms of the superimposed distortions. This confirmed the decisive role of the end wall effects on the quality of acquired results. The undistorted interferograms of the inlet flow in the region of the shock structure are needed to determine the actual angle of attack of the incoming flow onto the tested transonic cascade. Based on the presented results it is suggested for the future testing of flat cascades to modify the front part of the blades by appropriate side cut-offs to eliminate interferogram distortions.