diff --git a/pipeline/feature_extractors.py b/pipeline/feature_extractors.py
index cf9dbdb..38063de 100644
--- a/pipeline/feature_extractors.py
+++ b/pipeline/feature_extractors.py
@@ -47,6 +47,9 @@ class LaughterFeatureExtractor(FeatureExtractor):
 
     See: https://github.com/jrgillick/laughter-detection for the laughter-detection library
     """
+    _PREPEND_TIME = 7.0 # seconds before the laugh
+    _APPEND_TIME = 3.0 # seconds after the laugh
+
 
     def __init__(self, input_files=None, config=None):
         """It is expected that input_files is a SourceMedia object"""
@@ -94,13 +97,10 @@ class LaughterFeatureExtractor(FeatureExtractor):
         TODO: config for length adjustments per design doc
         TODO: play with numbers more to see what works best
         """
-        PREPEND = 7.0
-        APPEND = 3.0
-
         for feature in self.features:
             # do the pre & post adjustment
-            feature.interval.move_start(-PREPEND, relative=True)
-            feature.interval.move_end(APPEND, relative=True)
+            feature.interval.move_start(-self._PREPEND_TIME, relative=True)
+            feature.interval.move_end(self._APPEND_TIME, relative=True)
 
     def setup(self):
         """Setup the laughter feature extractor -- validate input files & config