Adjust delay of second sentence for process_ambi_items.py given extended output from reverb.

    else:

        ambi_format = "FOA"

    len_s1 = 0

    y = audio.SceneBasedAudio(ambi_format)

    for i in range(N_sources):

        # parse parameters from the scene description

            fs=cfg.IR_fs,

        )

        if i == 0:

            len_s1 = x.audio.shape[0]

        # convolve with the FOA/HOA2 IR

        if ambi_format == "FOA":

            x = reverb_foa(x, IR)

        # shift the second (and all other) source files (positive shift creates overlap, negative shift creates a gap)

        if i > 0:

            # get the length of the first source file

            N_delay = len(y.audio[:, 0])

            N_delay = len_s1

            # add the shift

            N_delay += int(-source_overlap * x.fs)