diff --git a/6_extract_impulse.py b/6_extract_impulse.py
new file mode 100644
index 0000000..d0db9c4
--- /dev/null
+++ b/6_extract_impulse.py
@@ -0,0 +1,104 @@
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+from scipy.interpolate import interp1d
+from config import Field_data_test_Config as cfg
+from config import Path_Config as pcfg
+
+# taper cut
+def apply_taper_window(data, start_idx, end_idx, left_frac=0.15, right_frac=0.15):
+    """
+    left_frac/right_frac: taper ratio (0–1)
+    """
+    n = len(data)
+    start = int(max(0, min(n, start_idx)))
+    end   = int(max(0, min(n, end_idx)))
+    if end <= start:
+        raise ValueError("end_idx must be greater than start_idx.")
+
+    N = end - start
+    L = int(np.clip(int(N * float(left_frac)),  0, N))
+    R = int(np.clip(int(N * float(right_frac)), 0, N - L))
+
+    win = np.ones(N, dtype=float)
+    if L > 0:
+        win[:L] = 0.5 - 0.5 * np.cos(np.linspace(0, np.pi, L))          # 0 -> 1
+    if R > 0:
+        win[-R:] = 0.5 - 0.5 * np.cos(np.linspace(np.pi, 0, R))         # 1 -> 0
+
+    mask = np.zeros(n, dtype=float)
+    mask[start:end] = win
+    return data * mask
+
+def shift_data_to_end(data, n):
+    if n < 0 or n > len(data):
+        raise ValueError("Shift length n must be between 0 and the data length.")
+    return np.concatenate((data[n:], data[:n]))
+
+def read_file(file_path, idx):
+    if file_path.endswith('.csv'):
+        data = pd.read_csv(file_path, header=None)
+        one_d_data = data.iloc[1:, idx].values
+    elif file_path.endswith('.txt'):
+        data = np.loadtxt(file_path)
+        one_d_data = data if data.ndim == 1 else data[:, 0]
+    else:
+        raise ValueError("Unsupported file format. Please provide a .csv or .txt file.")
+    return one_d_data
+
+def normalize_data(data):
+    max_abs_value = np.max(np.abs(data))
+    return data if max_abs_value == 0 else data / max_abs_value
+
+def interpolate_data(data, target_length=1000):
+    original_length = len(data)
+    x_original = np.linspace(0, 1, original_length)
+    x_target = np.linspace(0, 1, target_length)
+    interpolator = interp1d(x_original, data, kind='cubic')
+    return interpolator(x_target)
+
+def save_to_csv(data, output_path):
+    pd.DataFrame(data).to_csv(output_path, index=False, header=False)
+
+def plot_data(original_data, interpolated_data):
+    plt.figure(figsize=(10, 5))
+    plt.plot(original_data, label='Original (after taper)', linestyle='-')
+    plt.plot(
+        np.linspace(0, len(original_data)-1, len(interpolated_data)),
+        interpolated_data, label='Interpolated', linewidth=2
+    )
+    plt.legend(); plt.xlabel('Index'); plt.ylabel('Value')
+    plt.title('Soft-gated (tapered) Window + Interpolation')
+    plt.grid(alpha=0.3)
+    plt.show()
+
+# —— Main pipeline update: use a soft taper instead of a hard cut ——
+if __name__ == "__main__":
+    input_file = pcfg.PROCESSED_TEST_FILE
+    output_csv = pcfg.field_impulse
+    idx = cfg.refer_wave_idx
+    static_time = cfg.static_time
+    wavelet_range = cfg.wavelet_range  # [start_idx, end_idx]
+
+    # Optional: allow taper ratios to be configured in cfg; fall back to defaults here
+    taper_left_frac  = getattr(cfg, 'taper_left_frac',  0.15)  # 15% on the left
+    taper_right_frac = getattr(cfg, 'taper_right_frac', 0.15)  # 15% on the right
+
+    data = read_file(input_file, idx)
+    normalized_data = -normalize_data(data)
+
+    # Key step: apply a soft-gated window (taper at both ends) to the selected interval; samples outside become 0
+    tapered_data = apply_taper_window(
+        normalized_data,
+        wavelet_range[0], wavelet_range[1],
+        left_frac=taper_left_frac, right_frac=taper_right_frac
+    )
+
+    save_to_csv(tapered_data, './impulse/initial_impulse.csv')
+
+    # Then interpolate to the target length (recommended: interpolate after tapering)
+    interpolated_data = interpolate_data(tapered_data, cfg.time_window_length)
+    interpolated_data = shift_data_to_end(interpolated_data, static_time)
+
+    save_to_csv(interpolated_data, output_csv)
+    plot_data(tapered_data, interpolated_data)