Question about inverting blasting signals in meter-scale mining environment

[CSUPZW]

Dear sir, I have been trying to invert moment tensors of blasting signals with the mttime package, but I run into some problems when fitting the observed and synthesized waveforms （only contain Z component）.
Since distances between blasting sources and receivers is around 60 meters, the high-frequency portion of the blast signal is relatively large, as shown below.

The corner frequency reaches 161.9048 Hz, and it poses challenges for selecting parameters of the bandpass filter.
When the parameters are set [30Hz, 160Hz] to contain low and high frequency of blasting signals, the differences between observed and synthesized waveforms are too large.

When the parameters are set [120Hz,160Hz] containing only high frequency, the obtained results do not correspond to the characteristics of the blast signals.
Therefore, I would like to ask how to determine the filtering parameters applicable to the blast signal.
In addition, I found that the beach ball pattern of luna in the mttime package is inconsistent with the beach ball pattern in obspy.

Thank you for considering my question. I am eagerly anticipating your response.
Best regards,
Zhongwei Pei