diff --git a/source/_integrations/filter.markdown b/source/_integrations/filter.markdown
index d8f3af58bb2..e8e4116d644 100644
--- a/source/_integrations/filter.markdown
+++ b/source/_integrations/filter.markdown
@@ -78,16 +78,16 @@ filters:
       description: Algorithm to be used to filter data. Available filters are  `lowpass`, `outlier`, `range`, `throttle`, `time_throttle` and `time_simple_moving_average`.
       required: true
       type: string
+    precision:
+      description: Defines the precision of the filtered state, through the argument of round().
+      required: false
+      type: integer
+      default: 2
     window_size:
       description: Size of the window of previous states. Time based filters such as `time_simple_moving_average` will require a time period (size in time), while other filters such as `outlier` will require an integer (size in number of states). Time periods are in _hh:mm_ format and must be quoted.
       required: false
       type: [integer, time]
       default: 1
-    precision:
-      description: See [_lowpass_](#low-pass) filter. Defines the precision of the filtered state, through the argument of round().
-      required: false
-      type: integer
-      default: None
     time_constant:
       description: See [_lowpass_](#low-pass) filter. Loosely relates to the amount of time it takes for a state to influence the output.
       required: false
@@ -129,8 +129,6 @@ A = 1.0 - B
 LowPass(state) = A * previous_state + B * state
 ```
 
-The returned value is rounded to the number of decimals defined in (`precision`).
-
 ### Outlier
 
 The Outlier filter (`outlier`) is a basic Band-pass filter, as it cuts out any value outside a specific range.