diff --git a/gnommo/renderer.py b/gnommo/renderer.py
index b43bbfc..03caa9d 100644
--- a/gnommo/renderer.py
+++ b/gnommo/renderer.py
@@ -1292,17 +1292,20 @@ def build_filter_complex(
             _, first_video_source, _ = plan.narration_videos[0]
             use_channels = first_video_source.use_audio_channels
             if use_channels == "auto":
-                # Probe the real narration audio to pick left/right/both. In concat
-                # mode that's the first processed segment; otherwise the single
-                # narration input. (plan carries the dirs — this function doesn't.)
+                # Concat mode = pre-processed segments: preprocess already resolved
+                # the channel and baked it into the audio, so play it as-is ("both").
+                # Re-probing here means a full-file volumedetect pass on the narration
+                # (tens of seconds on a long take) for no benefit — that was the render
+                # command taking ~30s to build.
                 if narration_concat and plan.narration_segments:
-                    narration_path = plan.narration_segments[0].source_path
+                    use_channels = "both"
                 else:
+                    # Legacy single-input narration (not preprocessed): resolve once.
                     narration_path = _resolve_video_path(
                         plan.videos_dir, first_video_source,
                         plan.shared_assets_dir, plan.project_path,
                     )
-                use_channels = _resolve_auto_channel(narration_path)
+                    use_channels = _resolve_auto_channel(narration_path)
             channel_filter = _build_audio_channel_filter(use_channels)
             narration_volume = first_video_source.volume