[MSNoise] bugs in AC or its stack in version 1.6?

Dear Thomas, It is nice that the calculation of AC or SC within same station is considered using new parameters 'components_to_compute_single_station' in the new MSnoise. When i use the code, i find using AC always leads to a bad dvv time series. There is nothing related to an earthquake, but I can obtain clear coseismic change via SC. I check the output at each step and i find that the 'REF' waveform for AC has very low SNR (that is, coda of 'REF' with strong fluctuation) than the case of SC. In addition, the amplitude value of 'REF' waveform is very large, sometimes several orders larger than the 'MOV' stacks. Is this because the normalization of AC (or CC) waveform is removed in the version 1.6? Looking forward to hearing from you. Best wishes, Zhikun