diff --git a/render.py b/render.py
index 3352ea3..f5e1499 100644
--- a/render.py
+++ b/render.py
@@ -210,7 +210,6 @@ def compose(clips, fragment, target=150, base=1024, voice_over=None, options=Non
     scene_duration = int(get_scene_duration(scene) * fps)
     voice_overs = []
     sub_offset = 0
-    vo_min = 0
     subs = []
     print("--")
     print("Voice Over:")
@@ -221,6 +220,7 @@ def compose(clips, fragment, target=150, base=1024, voice_over=None, options=Non
                 gap = (5 * fps + random_int(seq, 10 * fps)) / fps
             else:
                 gap = (2 * fps + random_int(seq, 5 * fps)) / fps
+            gap = format_duration(gps, fps)
             if int((sub_offset + gap)* fps) > scene_duration:
                 gap = format_duration((scene_duration - int(sub_offset * fps)) / fps, fps)
             scene['audio-center']['A1'].append({