diff --git a/tests/test_reference_runs/test_pre_deployment.py b/tests/test_reference_runs/test_pre_deployment.py
index 0bb880184908e03f815e52b31b98c28509bbe18b..a04be816ef2c7eee6b8b071bb12ab82062917340 100644
--- a/tests/test_reference_runs/test_pre_deployment.py
+++ b/tests/test_reference_runs/test_pre_deployment.py
@@ -48,7 +48,7 @@ def validate_h5files(
 
     Args:
         f: file to validate
-        ref_folder (_type_): Reference folder which has the same filename
+        ref_folder: Reference folder which has the same filename
             as of the file to validate.
 
     Returns:
@@ -96,13 +96,13 @@ def validate_constant_file(
         h5_copy_except.h5_copy_except_paths(
             sfile,
             hp1,
-            ["report", "creation_time"],
+            ["report"],
         )
     with h5py.File(ref_folder / test_file.name, "r") as sfile:
         h5_copy_except.h5_copy_except_paths(
             sfile,
             hp2,
-            ["report", "creation_time"],
+            ["report"],
         )
     hp1.close()
     hp2.close()
@@ -152,7 +152,25 @@ def test_xfel_calibrate(
     test_key: str, val_dict: dict, release_test_config: Tuple[bool, bool, bool, bool]
 ):
     """Test xfel calibrate detectors and calibrations written
-    in the given callab_test YAML file.
+    in the given callab_test.py file.
+
+    xfel-calibrate CL is ran over each test_key. Expected a val_dict of
+    1. out-folder, 2. in-folder, 3. run numbers, 4.karabo-id, 5. detector type
+    6. calibration type, and 7. reference folder to compare the data with.
+    The val_dict can/should acquire the needed custom configurations for each test.
+    e.g. AGIPD notebooks needs to differentiate between the different `karabo-id-control`
+    which differ across instruments.
+
+    Each calibration test is ran sequentially. Each calibration test consists of multiple tests.
+    1. Testing if all SLURM jobs were successfull.
+    2. Testing if there is an available report.
+    3. Validate number of H5 files against files in reference folder.
+    4. Validate numberical values for the H5 files against reference files.
+      a. CORRECTED files are compared using MD5 checksum.
+      b. CONSTANTS local files saved using cal_tools are the one being tested.
+      Unfortunately they can't be compared using MD5 checksum. As the metadata in files
+      consists of the report path. which differ between each test. So the metadata is cleaned
+      from the report path first before comparing the H5 constant files numerically.
     Args:
         test_key : Key for the xfel-calibrate test.
         val_dict: Dictionary of the configurations for the running test.