Since February 2019, the InSight NASA mission and its seismometer, the SEIS, allows us to study and analyse the motion of the martian ground. In order to investigate the reliability of all event detections made during this mission and so, the possible amount of available sources to be further inverted for structure, we developed a new approach to quantify the equirepartition of the seismic signal. This statistic method allows to point out low energy transient signals that are characterized by asymmetric distributions of amplitudes in the raw signal. At the moment we are able to identify artifacts as glitches (instrumental response to an external perturbation) and the seismic signatures of atmospheric pressure drops that are travelling close to the lander (also visible on the pressure sensor). Although this method can be used 1) to classify the different kind of glitches and/or 2) to clean the raw signal, before any deconvolution, we think it can be well-suited 3) to detect low-energy-long-period seismic signals. This approach can be used on Earth data to better characterize the continuous seismic signal recorded at permanent stations.