diff --git a/README.md b/README.md index aa14434..956263e 100644 --- a/README.md +++ b/README.md @@ -524,7 +524,7 @@ NumberListGenerator 節點可根據自訂參數創建數字列表，支援有序 ### 功能特色 -**分割音檔為清單**：將聲音檔分割為清單，長篇數字人時可分段採樣。 +**分割音檔為清單**：將聲音檔分割為清單，長篇數字人時可分段採樣，目前已經更新音檔處理的邏輯，增加FADE功能 **必需輸入:** - **videofps / 畫格** (Float): 每秒多少格 diff --git a/Readme/demo5.jpg b/Readme/demo5.jpg index 878b912..3890c30 100644 Binary files a/Readme/demo5.jpg and b/Readme/demo5.jpg differ diff --git a/nodes.py b/nodes.py index b26a5f2..b6ba2f4 100644 --- a/nodes.py +++ b/nodes.py @@ -34,6 +34,10 @@ class AudioListGenerator: "videofps": ("FLOAT", {"default": 23.976, "min": 1.0, "step": 0.001}), "samplefps": ("INT", {"default": 81, "min": 1}), "pad_last_segment": ("BOOLEAN", {"default": True}), + }, + "optional": { + "crossfade_duration": ("FLOAT", {"default": 0.1, "min": 0.0, "max": 2.0, "step": 0.01}), + "crossfade_type": (["linear", "cosine", "equal_power"], {"default": "cosine"}), } } @@ -43,27 +47,50 @@ class AudioListGenerator: FUNCTION = "split" CATEGORY = "ListHelper" - def split(self, waveform, videofps, samplefps, pad_last_segment): + def split(self, waveform, videofps, samplefps, pad_last_segment, crossfade_duration=0.1, crossfade_type="cosine"): audio_tensor = waveform["waveform"] # shape: [1, C, N] sample_rate = waveform["sample_rate"] total_samples = audio_tensor.shape[-1] segment_duration_seconds = samplefps / videofps samples_per_segment = int(segment_duration_seconds * sample_rate) + crossfade_samples = int(crossfade_duration * sample_rate) audio_list = [] + # 確保交叉淡化時間不會超過段落長度的一半 + crossfade_samples = min(crossfade_samples, samples_per_segment // 2) + for i in range(0, total_samples, samples_per_segment): end_idx = min(i + samples_per_segment, total_samples) - segment = audio_tensor[:, :, i:end_idx].clone() + + # 計算實際的開始和結束位置，考慮交叉淡化 + actual_start = max(0, i - crossfade_samples) if i > 0 else 0 + actual_end = min(total_samples, end_idx + crossfade_samples) if end_idx < total_samples else end_idx + + # 提取包含交叉淡化部分的音頻段 + extended_segment = audio_tensor[:, :, actual_start:actual_end].clone() + + # 應用交叉淡化效果 + if crossfade_samples > 0: + extended_segment = self._apply_crossfade( + extended_segment, + crossfade_samples, + crossfade_type, + actual_start, + i, + end_idx, + actual_end + ) - segment_len = segment.shape[-1] - if pad_last_segment and segment_len < samples_per_segment: + # 如果需要填充最後一個段落 + segment_len = extended_segment.shape[-1] + if pad_last_segment and end_idx == total_samples and segment_len < samples_per_segment: pad_len = samples_per_segment - segment_len - segment = F.pad(segment, (0, pad_len)) + extended_segment = F.pad(extended_segment, (0, pad_len)) audio_obj = { - "waveform": segment, + "waveform": extended_segment, "sample_rate": sample_rate } @@ -71,6 +98,85 @@ class AudioListGenerator: return len(audio_list), audio_list + def _apply_crossfade(self, segment, crossfade_samples, crossfade_type, actual_start, segment_start, segment_end, actual_end): + """ + 對音頻段應用交叉淡化效果 + + Args: + segment: 音頻段張量 [1, C, T] + crossfade_samples: 交叉淡化的樣本數 + crossfade_type: 交叉淡化類型 + actual_start: 實際開始位置 + segment_start: 段落開始位置 + segment_end: 段落結束位置 + actual_end: 實際結束位置 + """ + if crossfade_samples == 0: + return segment + + segment_length = segment.shape[-1] + + # 創建淡化曲線 + fade_curve = self._create_fade_curve(crossfade_samples, crossfade_type) + + # 應用淡入效果（段落開始處） + if actual_start < segment_start: + fade_in_length = min(crossfade_samples, segment_length) + fade_in_curve = fade_curve[:fade_in_length] + + # 擴展維度以匹配音頻張量 [1, C, fade_in_length] + fade_in_curve = fade_in_curve.unsqueeze(0).unsqueeze(0) + fade_in_curve = fade_in_curve.expand(segment.shape[0], segment.shape[1], -1) + + segment[:, :, :fade_in_length] *= fade_in_curve + + # 應用淡出效果（段落結束處） + if actual_end > segment_end: + fade_out_length = min(crossfade_samples, segment_length) + fade_out_curve = fade_curve[:fade_out_length].flip(0) # 反轉淡化曲線 + + # 擴展維度以匹配音頻張量 + fade_out_curve = fade_out_curve.unsqueeze(0).unsqueeze(0) + fade_out_curve = fade_out_curve.expand(segment.shape[0], segment.shape[1], -1) + + segment[:, :, -fade_out_length:] *= fade_out_curve + + return segment + + def _create_fade_curve(self, length, fade_type): + """ + 創建淡化曲線 + + Args: + length: 淡化長度（樣本數） + fade_type: 淡化類型 ("linear", "cosine", "equal_power") + + Returns: + 淡化曲線張量 + """ + import torch + import math + + if fade_type == "linear": + # 線性淡化：從0到1 + curve = torch.linspace(0.0, 1.0, length) + + elif fade_type == "cosine": + # 餘弦淡化：更平滑的過渡 + t = torch.linspace(0.0, math.pi/2, length) + curve = torch.sin(t) + + elif fade_type == "equal_power": + # 等功率淡化：保持總功率恆定 + t = torch.linspace(0.0, math.pi/2, length) + curve = torch.sin(t) + + else: + # 預設使用線性淡化 + curve = torch.linspace(0.0, 1.0, length) + + return curve + class AudioToFrameCount: @classmethod def INPUT_TYPES(cls): @@ -812,6 +918,176 @@ class SaveVideoPath: ) return (full_path,) + +class TimestampToLrcNode: + """ + ComfyUI節點：將時間戳格式轉換為LRC歌詞格式 + """ + + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "input_text": ("STRING", { + "multiline": True, + "default": "" + }), + } + } + + RETURN_TYPES = ("STRING",) + RETURN_NAMES = ("lrc_output",) + FUNCTION = "convert_to_lrc" + CATEGORY = "text/processing" + + def convert_to_lrc(self, input_text): + """ + 將時間戳格式轉換為LRC格式 + 輸入格式: >> 0:00-0:04\n>> 文本內容 + 輸出格式: [00:00.00]文本內容 + """ + + lines = input_text.strip().split('\n') + lrc_lines = [] + current_time = None + current_text = "" + + for line in lines: + line = line.strip() + + # 檢查是否為時間戳行 (格式: >> 0:00-0:04) + time_match = re.match(r'^>>\s*(\d+):(\d+)-(\d+):(\d+)$', line) + if time_match: + # 如果之前有累積的文本，先處理它 + if current_time is not None and current_text.strip(): + lrc_lines.append(f"[{current_time}]{current_text.strip()}") + + # 解析開始時間 + start_min = int(time_match.group(1)) + start_sec = int(time_match.group(2)) + current_time = f"{start_min:02d}:{start_sec:02d}.00" + current_text = "" + + # 檢查是否為文本行 (格式: >> 文本內容) + elif line.startswith('>> '): + text_content = line[3:].strip() # 移除 ">> " 前綴 + if text_content: # 只添加非空文本 + if current_text: + current_text += " " + text_content + else: + current_text = text_content + + # 處理空行或其他格式 + elif line == '' or line == '>>': + # 空行保持當前狀態，不做處理 + continue + else: + # 其他格式的行，嘗試作為文本處理 + if line and current_time is not None: + if current_text: + current_text += " " + line + else: + current_text = line + + # 處理最後一段文本 + if current_time is not None and current_text.strip(): + lrc_lines.append(f"[{current_time}]{current_text.strip()}") + + # 合併結果 + lrc_output = '\n'.join(lrc_lines) + + return (lrc_output,) + +try: + import opencc +except ImportError: + print("請安裝opencc庫: pip install opencc-python-reimplemented") + opencc = None + +class ChineseConverterNode: + """ + ComfyUI節點：中文簡繁轉換 + 使用opencc庫進行高質量轉換 + 布林開關控制：True=簡體轉繁體，False=繁體轉簡體 + """ + + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "input_text": ("STRING", { + "multiline": True, + "default": "" + }), + "simp_to_trad": ("BOOLEAN", { + "default": True, + "label_on": "簡體→繁體", + "label_off": "繁體→簡體" + }), + } + } + + RETURN_TYPES = ("STRING",) + RETURN_NAMES = ("converted_text",) + FUNCTION = "convert_chinese" + CATEGORY = "text/processing" + + def __init__(self): + """初始化轉換器""" + if opencc is None: + self.s2t_converter = None + self.t2s_converter = None + print("錯誤：opencc庫未安裝，請執行: pip install opencc-python-reimplemented") + else: + try: + # 簡體轉繁體轉換器 + self.s2t_converter = opencc.OpenCC('s2t.json') + # 繁體轉簡體轉換器 + self.t2s_converter = opencc.OpenCC('t2s.json') + except Exception as e: + print(f"opencc初始化失敗: {e}") + self.s2t_converter = None + self.t2s_converter = None + + def convert_chinese(self, input_text, simp_to_trad): + """ + 轉換中文文本 + + Args: + input_text: 輸入文本 + simp_to_trad: True=簡體轉繁體，False=繁體轉簡體 + + Returns: + 轉換後的文本 + """ + + if not input_text.strip(): + return ("",) + + # 檢查opencc是否可用 + if opencc is None: + error_msg = "錯誤：請先安裝opencc庫\n執行命令: pip install opencc-python-reimplemented" + print(error_msg) + return (error_msg,) + + try: + if simp_to_trad: + # 簡體轉繁體 + if self.s2t_converter is None: + self.s2t_converter = opencc.OpenCC('s2t.json') + converted_text = self.s2t_converter.convert(input_text) + else: + # 繁體轉簡體 + if self.t2s_converter is None: + self.t2s_converter = opencc.OpenCC('t2s.json') + converted_text = self.t2s_converter.convert(input_text) + + return (converted_text,) + + except Exception as e: + error_msg = f"轉換失敗: {str(e)}" + print(error_msg) + return (error_msg,) @@ -825,6 +1101,8 @@ NODE_CLASS_MAPPINGS = { "CeilDivide": CeilDivide, "LoadVideoPath": LoadVideoPath, "SaveVideoPath": SaveVideoPath, + "TimestampToLrcNode": TimestampToLrcNode, + "ChineseConverterNode": ChineseConverterNode, } NODE_DISPLAY_NAME_MAPPINGS = { @@ -837,5 +1115,7 @@ NODE_DISPLAY_NAME_MAPPINGS = { "CeilDivide": "CeilDivide", "LoadVideoPath": "LoadVideoPath", "SaveVideoPath": "SaveVideoPath", + "TimestampToLrcNode": "TimestampToLrcNode", + "ChineseConverterNode": "ChineseConverterNode", }