remove squeeze from dsp

[hoyer-a]

Remove squeeze from dsp.find_impulse_response_maximum() to match updated behavior in pyfar.

[ahms5]

Thank you for taking care! Maybe we should update the docstring about what is happening.

[hoyer-a]

Like a more detailed explanation what the function does?

The docs in general could use a makeover. In some cases, and also here, the docstring calls for ndarrays as input parameter, but the code clearly uses pyfar Signals.

[ahms5]

oh yes, now I see.
maybe sth like this:

    """Find the maximum of an impulse response as argmax(h(t)).
    Performs an initial SNR check according to a defined threshold level in dB.

    Parameters
    ----------
    impulse_response : pyfar.Signal
        The impulse response
    threshold : double, optional
        Threshold SNR value in dB. A warning is raised if the threshold is not reached. By default 20dB.

    Returns
    -------
    max_sample : np.ndarray of int
        Sample at which the impulse response starts for each channel. The shape will be the ``cshape`` of  ``impulse_response``

    Note
    ----
    The function tries to estimate the SNR in the IR based on the signal energy
in the last 10 percent of the IR.

another point is, is pyfar.dsp.find_impulse_response_start is this function doing the same?

[hoyer-a]

Looks like the calculation is exactly the same, we could just use the pyfar function.
That already happens in pyrato's find_impulse_response_start().

There might be even more cases, i can take a look at it.