From 96fce4e28d2501673fc0d60f8da808d290a50b50 Mon Sep 17 00:00:00 2001 From: Hawk Lee Date: Sun, 15 Feb 2026 20:05:45 +0800 Subject: [PATCH] fix(stitcher): prevent repeated audio from padding-induced segment overlap MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Track each speaker's previous cut_end. After applying padding, clamp cut_start to never be earlier than the speaker's previous cut_end. This eliminates the 200ms overlap that caused A's tail audio to replay when switching A→B→A. --- aiia_podcast_stitcher.py | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/aiia_podcast_stitcher.py b/aiia_podcast_stitcher.py index 14dea28..c055a12 100644 --- a/aiia_podcast_stitcher.py +++ b/aiia_podcast_stitcher.py @@ -425,6 +425,8 @@ class AIIA_Podcast_Stitcher: idx_A = 0 idx_B = 0 prev_speaker = None + # 跟踪每个说话人上一个片段的实际 cut_end，防止 padding 导致重叠 + prev_cut_end = {"A": 0.0, "B": 0.0} for item in map_items: if item.get("type") == "pause": @@ -480,6 +482,11 @@ class AIIA_Podcast_Stitcher: cut_start = max(0, cut_start - padding) cut_end = min(len(wav) / sr, cut_end + padding) + # 防重叠：确保 cut_start 不早于同一说话人上一个片段的 cut_end + if cut_start < prev_cut_end[speaker]: + cut_start = prev_cut_end[speaker] + prev_cut_end[speaker] = cut_end + start_sample = int(cut_start * sr) end_sample = int(cut_end * sr) end_sample = min(end_sample, len(wav))