Tobias Megies

I had a good look at this now, and I dont see a reason why the decimation info is not stored with base type ResponseStage.. I think it's an overlook...

fix is in main branches and will be in upcoming releases

Yeah we need to change this somehow. Side note: always a good idea to do a `st.merge(-1)` before any processing, also you should always use like 2 minutes at start/end...

I can confirm the problem with your code and minimal SDS file tree. Removing both files `*.2019.032` (which do not contain any requested data) still shows the issue for me,...

> You should also remove the IRIS examples that no longer work, e.g., tutorial/code_snippets/seismometer_correction_simulation_3.py Fixed this one, see 57b0405c47812f80. I didn't see any other use of obspy.iris in tutorial.. We...

Here's an example use case: ```python from math import cos, sin, radians import numpy as np import matplotlib.pyplot as plt from obspy.signal.invsim import optimize_paz, _paz_to_freq_resp frequencies = np.array( [0.5, 0.8,...

> We've been doing the same @dsentinel would be good to see the code, no need to use the "real" values if it might be problematic, you can try it...

> We tend to force the sign of the poles. However, we are starting with an initial model and then perturbing that. Do you think adding an optional `initial_guess` would...

Thanks for linking your PZcalc project. How would i call it given the F-A-P list from the example used here? https://github.com/obspy/obspy/pull/2618#issuecomment-633974011

@tychoaussie thanks, seems pretty similar. The main difference I see is that you have quite strong constraints on the poles/zeros, which might be a good way to go in most...