In tide gauge records, the separation of tidal from non-tidal contents is in general done by specialists after applying high pass or bandpass filtering techniques. However, there is always a tradeoff between losing some valuable contents of the signal while trying to remove the tide. In this manuscript, we deal with two significant tide gauge records of the 1945 Makran tsunami. The importance of these data is well known for researchers since these are the only valid tide gauge data for any evaluation of tsunami in the western Makran region. Wavelet decomposition technique was applied as a tool by us to minimize the influence of tidal contents on these data. Several parameters were set and applied to several types of wavelets in various temporal series and frequency ranges. The goal was to find the most appropriate wavelet parameter and type, to distinguish tsunami data from tidal contents. Our results show that the application of wavelets in comparison to conventional methods yields to lesser RMS errors before the arrival of the tsunami wave and therefore produces minor data loss after applying. Moreover, after applying suitable wavelets to data, we have noticed a sudden decrease in the sea level before the arrival of tsunami waves, which is consistent with eyewitness reports. The result of our work is beneficial for further evaluation of tsunami modeling in western Makran. The procedure introduced can be applied at any tide gauge data for tidal analysis with wavelet decomposition techniques.