diff --git a/docs/changelog.md b/docs/changelog.md
index 20668730..c134b7a6 100644
--- a/docs/changelog.md
+++ b/docs/changelog.md
@@ -95,6 +95,8 @@ Fixed:
 - Fixed counting threshold for `CookieboxCalibration` (!498).
 - [CookieboxCalibration][extra.applications.CookieboxCalibration] bug fix for
   new `extra.applications.base.SerializableMixin` interface (!506).
+- [CookieboxCalibration][extra.applications.CookieboxCalibration] transmission
+  corrected based on pulse energy (!512).
 
 Changed:
 
diff --git a/src/extra/applications/cookiebox.py b/src/extra/applications/cookiebox.py
index 522ee090..0c3f75b2 100644
--- a/src/extra/applications/cookiebox.py
+++ b/src/extra/applications/cookiebox.py
@@ -13,6 +13,7 @@
 import xarray as xr
 import pandas as pd
 from scipy.signal import find_peaks
+from scipy.ndimage import gaussian_filter
 
 from .base import SerializableMixin
 from .cookiebox_deconvolve import TOFAnalogResponse
@@ -47,7 +48,7 @@ def search_offset(trace: np.ndarray, sigma: float=20) -> int:
     peak_idx = np.argmax(smoothened)
     return peak_idx - 200
 
-def search_roi(roi: np.ndarray) -> np.ndarray:
+def search_roi(roi: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
     """
     Find highest peaks in the 1D trace.
 
@@ -56,9 +57,21 @@ def search_roi(roi: np.ndarray) -> np.ndarray:
 
     Returns: Peak position.
     """
-    import scipy
-    p, _ = scipy.signal.find_peaks(roi, prominence=(0.25*np.max(roi), None))
-    return p
+    roi_smooth = gaussian_filter(roi, sigma=5)
+    p, prop = find_peaks(roi_smooth, prominence=0.25*np.max(roi_smooth), width=1, height=0)
+    idx = np.argsort(prop["peak_heights"])[::-1]
+    p = p[idx]
+    w = prop["widths"][idx]
+    if len(p) == 0:
+        logging.info(f"Failed to find peaks.")
+        return -1, 0
+    if len(p) == 1:
+        logging.info(f"Failed to find two peaks. If there is no Auger peak, set start_roi=0")
+        return -1, 0
+    idx = np.argsort(p[:2])
+    p = p[:2][idx]
+    w = w[:2][idx]
+    return p[-1], w[-1]
 
 def model(ts: np.ndarray, c: float, e0: float, t0: float) -> np.ndarray:
     """
@@ -177,16 +190,13 @@ def calc_mean(itr: Tuple[int, int], scan: Scan, xgm_data: xr.DataArray, tof: Dic
         tof_data = tof_data.pulse_data(pulse_dim='pulseIndex', parallel=parallel)
 
         # select XGM
-        if xgm_threshold > 0:
-            mask = xgm_data.coords["trainId"].isin(good_ids)
-            sel_xgm_data = xgm_data[mask]
-            tof_data = tof_data.loc[sel_xgm_data > xgm_threshold, :]
-            tof_xgm_data = sel_xgm_data.loc[sel_xgm_data > xgm_threshold]
+        mask = xgm_data.coords["trainId"].isin(good_ids)
+        sel_xgm_data = xgm_data[mask]
+        tof_data = tof_data.loc[sel_xgm_data > xgm_threshold, :]
+        tof_xgm_data = sel_xgm_data.loc[sel_xgm_data > xgm_threshold].to_numpy()
 
-        out_data = -tof_data.mean('pulse')
-        out_xgm = 0.0
-        if xgm_threshold > 0:
-            out_xgm = tof_xgm_data.mean('pulse').to_numpy()
+        out_data = -tof_data.mean("pulse")
+        out_xgm = np.mean(tof_xgm_data)
     else:
         # option 2: count photon peaks
         # in this case, ignore the XGM, as it is only used for cleaning the data
@@ -198,9 +208,7 @@ def calc_mean(itr: Tuple[int, int], scan: Scan, xgm_data: xr.DataArray, tof: Dic
         out_data, _ = np.histogram(tof_data.edge, bins=bins, weights=-tof_data.amplitude)
 
         out_data = xr.DataArray(out_data, dims=('sample'), coords={'sample': bins[:-1]})
-        out_xgm = 0.0
-        if xgm_threshold > 0:
-            out_xgm = xgm_data.mean('pulse').to_numpy()
+        out_xgm = np.mean(xgm_data.to_numpy())
 
     if correction_fn is not None:
         out_data = correction_fn[tof_id](out_data)
@@ -691,16 +699,13 @@ def update_roi(self, parallel=None):
         logging.info("Reading calibration data ... (this takes a while)")
         self.select_calibration_data(parallel)
         # find RoI if needed
-        if (self.auger_start_roi is None
-            or self.start_roi is None
-            or self.stop_roi is None):
-            logging.info("Finding RoI ...")
-            logging.info("(This may fail. If it does, please provide a `auger_start_roi`, `start_roi` and `stop_roi`.)")
-            for tof_id in self.kwargs_adq.keys():
+        for tof_id in self.kwargs_adq.keys():
+            if (self.auger_start_roi[tof_id] is None
+                or self.start_roi[tof_id] is None
+                or self.stop_roi[tof_id] is None):
+                logging.info(f"Finding RoI for TOF {tof_id}...")
                 self.find_roi(tof_id)
-            logging.info(f"Auger start RoIs found: {self.auger_start_roi}")
-            logging.info(f"Start RoIs found: {self.start_roi}")
-            logging.info(f"Stop RoIs found: {self.stop_roi}")
+                logging.info(f"Start RoIs found: {self.start_roi[tof_id]}")
 
     def update_fit_result(self):
         """
@@ -781,19 +786,17 @@ def find_roi(self, tof_id: int):
         Args:
           tof_id: The eTOF ID.
         """
-        auger = list()
-        roi = list()
+        p = list()
+        w = list()
+        N = self.calibration_data[tof_id].shape[-1]
         for e in range(self.calibration_data[tof_id].shape[0]):
             d = self.calibration_data[tof_id][e, :]
-            peaks = search_roi(d)
-            peaks = sorted(peaks)
-            if len(peaks) < 2:
-                logging.info(f"Failed to find peaks for eTOF {tof_id}, energy index {e}. "
-                             f"Check the data quality.")
+            peak, width = search_roi(d)
+            if peak < 0:
                 continue
-            auger += [peaks[0]]
-            roi += [peaks[1]]
-        if len(auger) == 0 or len(roi) == 0:
+            p += [peak]
+            w += [width]
+        if len(p) == 0:
             logging.info(f"No peaks found in eTOF {tof_id}. "
                          f"Check the data quality. "
                          f"I will set the RoI to collect non-sense,"
@@ -801,18 +804,14 @@ def find_roi(self, tof_id: int):
                          f"It will also be masked.")
             self.auger_start_roi[tof_id] = 0
             self.start_roi[tof_id] = 100
-            self.stop_roi[tof_id] = 200
+            self.stop_roi[tof_id] = N
             self.mask[tof_id] = False
             return
-        a = min(auger)
-        b = min(roi)
-        dab = abs(b - a)
-        stop_roi = max(roi) + int(dab/2)
-        auger_start_roi = int(a - dab/2)
-        start_roi = int(b - dab/2)
-        self.auger_start_roi[tof_id] = auger_start_roi
-        self.start_roi[tof_id] = start_roi
-        self.stop_roi[tof_id] = stop_roi
+        idx = np.argmin(p)
+        pos = p[idx] - 2*w[idx]
+        self.auger_start_roi[tof_id] = 0
+        self.start_roi[tof_id] = int(pos)
+        self.stop_roi[tof_id] = N
 
     def plot_calibration_data(self):
         """Plot data for checks.
@@ -966,9 +965,8 @@ def calculate_calibration_and_transmission(self, tof_id: int):
             detected = self.tof_fit_result[tof_id].A
             #produced = self.calibration_mean_xgm[tof_id] * self.tof_fit_result[tof_id].Aa * dsig_dth
             #produced = self.tof_fit_result[tof_id].Aa * dsig_dth
-            produced = dsig_dth
-            if produced == 0:
-                produced += 1e-1
+            produced = dsig_dth*self.calibration_mean_xgm[tof_id][mask][eidx]
+            produced[produced == 0] = 0.1 # avoid division by zero
             en = detected[mask][eidx]/produced
             # interpolate normalization
             self.normalization[tof_id] = np.interp(self.energy_axis,
diff --git a/tests/test_applications_cookiebox.py b/tests/test_applications_cookiebox.py
index 97646cba..2cb84233 100644
--- a/tests/test_applications_cookiebox.py
+++ b/tests/test_applications_cookiebox.py
@@ -254,6 +254,52 @@ def test_fit(tmp_path):
     for tof_id, v in correct.items():
         assert np.allclose(cal_read.model_params[tof_id], v, rtol=1e-2, atol=1e-2)
 
+def test_detect_roi_and_fit_single_channel(mock_sqs_etof_calibration_run, tmp_path, mock_etof_mono_energies, mock_etof_calibration_constants):
+    # use mock data
+    pulse_timing = 'SQS_RR_UTC/TSYS/TIMESERVER'
+    monochromator_energy = 'SA3_XTD10_MONO/MDL/PHOTON_ENERGY'
+    digitizer = 'SQS_DIGITIZER_UTC4/ADC/1:network'
+    digitizer_control = 'SQS_DIGITIZER_UTC4/ADC/1'
+    pulse_energy = 'SQS_DIAG1_XGMD/XGM/DOOCS'
+    mock_sqs_etof_calibration_run = mock_sqs_etof_calibration_run.select([pulse_timing,
+                                  digitizer, digitizer_control,
+                                  pulse_energy, f"{pulse_energy}:output",
+                                  monochromator_energy], require_all=True).select_trains(np.s_[10:])
+    channel_name = "1_A"
+    tof_ids = [0]
+    tof_channel = {}
+    tof_channel[0] = AdqRawChannel(mock_sqs_etof_calibration_run,
+                                   channel_name,
+                                   digitizer=digitizer,
+                                   first_pulse_offset=1000)
+    scan = Scan(mock_sqs_etof_calibration_run[monochromator_energy, "actualEnergy"], resolution=2)
+    energy_axis = np.linspace(965, 1070, 160)
+    xgm = XGM(mock_sqs_etof_calibration_run, pulse_energy)
+    cal = CookieboxCalibration(
+                    auger_start_roi=None,
+                    start_roi=None,
+                    stop_roi=None,
+    )
+    cal.setup(run=mock_sqs_etof_calibration_run, energy_axis=energy_axis, tof_settings=tof_channel,
+              xgm=xgm,
+              scan=scan)
+    correct_energies = np.unique(mock_etof_mono_energies)
+    correct_constants = np.array(mock_etof_calibration_constants)
+    for tof_id in tof_ids:
+        assert np.allclose(cal.tof_fit_result[tof_id].energy, correct_energies, rtol=1e-2, atol=1e-2)
+
+        energy = correct_energies
+
+        # get calibration curve
+        c, e0, t0 = cal.model_params[tof_id]
+        ts = t0 + np.sqrt(c/(energy - e0))
+
+        c_true, e0_true, t0_true = correct_constants
+        ts_true = t0_true + np.sqrt(c_true/(energy - e0_true))
+
+        # check how well it matches
+        assert np.allclose(ts, ts_true, rtol=1e-2, atol=1e-2)
+
 def test_avg_and_fit_single_channel(mock_sqs_etof_calibration_run, tmp_path, mock_etof_mono_energies, mock_etof_calibration_constants):
     # use mock data
     pulse_timing = 'SQS_RR_UTC/TSYS/TIMESERVER'