diff --git a/aeon/dj_pipeline/__init__.py b/aeon/dj_pipeline/__init__.py
index 3e4b3150..1225020a 100644
--- a/aeon/dj_pipeline/__init__.py
+++ b/aeon/dj_pipeline/__init__.py
@@ -40,7 +40,7 @@ def fetch_stream(query, drop_pk=True, round_microseconds=True):
     Args:
         query (datajoint.Query): A query object containing data from a Stream table
         drop_pk (bool, optional): Drop primary key columns. Defaults to True.
-        round_microseconds (bool, optional): Round timestamps to microseconds. Defaults to False.
+        round_microseconds (bool, optional): Round timestamps to microseconds. Defaults to True.
             (this is important as timestamps in mysql is only accurate to microseconds)
     """
     df = (query & "sample_count > 0").fetch(format="frame").reset_index()
diff --git a/aeon/dj_pipeline/analysis/block_analysis.py b/aeon/dj_pipeline/analysis/block_analysis.py
index 996500e1..21e30653 100644
--- a/aeon/dj_pipeline/analysis/block_analysis.py
+++ b/aeon/dj_pipeline/analysis/block_analysis.py
@@ -251,6 +251,7 @@ def make(self, key):
             if encoder_df.empty:
                 encoder_df["distance_travelled"] = 0
             else:
+                # -1 is for placement of magnetic encoder, where wheel movement actually decreases encoder
                 encoder_df["distance_travelled"] = -1 * analysis_utils.distancetravelled(encoder_df.angle)
                 encoder_df = encoder_df.resample(f"{freq}ms").first()
 
@@ -327,7 +328,7 @@ def make(self, key):
                 )
                 pos_df = fetch_stream(pos_query)[block_start:block_end]
                 pos_df["likelihood"] = np.nan
-                # keep only rows with area between 0 and 1000
+                # keep only rows with area between 0 and 1000 - likely artifacts otherwise
                 pos_df = pos_df[(pos_df.area > 0) & (pos_df.area < 1000)]
             else:
                 pos_query = (