Ethereal abstract shot. A dark, misty void representing the Bardo state. A single, small glowing orb of white light (representing a soul) floats aimlessly in the fog, looking lonely. In the far background, a piercing beam of warm golden light begins to cut through the darkness. Mystery, slight tension turning to hope. Soundtrack: Melancholic Erhu solo in D Major, low drone sound.DONA莊嚴的標準中文唱:將入中陰險境,輪轉生死,痛苦無依。生成16:9影頻
這裡有一個非常重要的技術觀念需要向您說明:VEO 3.0(以及目前的 Sora、Runway 等 AI 影片模型)主要強項在於「視覺生成」。
雖然 VEO 可以生成簡單的背景音效(如風聲、簡單的氛圍音),但它無法透過提示詞精準地「唱出」特定的歌詞(如:阿彌陀佛)、特定的旋律(D大調),或是精確演奏出多種樂器的合奏。
✅ 正確的製作流程是「影音分離,後製合一」:
聲音 (Audio): 使用之前提供的 Suno AI 提示詞,生成那首 100 秒的完整助念歌曲(這才是真正有歌詞、有月琴伴奏的音檔)。
畫面 (Visual): 使用 VEO 3.0 生成 DONA 「正在演唱和彈奏」的畫面。
合成 (Edit): 在剪輯軟體(VIDS)中,將 Suno 的音樂設為背景,把 VEO 的影片鋪在上面。
為了讓 VEO 生成的影片看起來**「像是在真唱」,我修改了以下 8 段提示詞。我移除了無效的 "Soundtrack" 指令,改為強化描述她的「嘴部動作 (Lip-syncing)」與「彈奏指法 (Strumming action)」**,這樣當您把 Suno 的音樂疊上去時,畫面會非常逼真。
請依序使用這些提示詞生成影片,生成時請選擇 "No Audio" (無聲模式),因為我們要用 Suno 的高品質音樂來取代。
SCENE 1:開場・聖境現前
重點: DONA 嘴型緩慢張合,神情莊嚴。
VEO 3.0 Prompt:
Cinematic wide shot, slow camera push in. DONA (18yo, blonde, white blazer, feather cuffs, AirPods Max) stands on a cloud platform. She is VISIBLY SINGING with mouth opening and closing slowly in a solemn manner, and strumming her pale maple Yueqin. Her feather cuffs blow in a divine breeze. Behind her, celestial clouds part to reveal a colossal, glowing Amitabha Buddha. Iridescent sky. 8k resolution.
SCENE 2:中陰救度
重點: 表情轉為悲憫,視線向下,嘴型像是在呼喚。
VEO 3.0 Prompt:
Ethereal abstract shot. DONA (18yo, blonde, white blazer) is SINGING with a compassionate, concerned expression, looking slightly downward. Her fingers are actively plucking the Yueqin strings. She is surrounded by a dark, misty void. A small glowing orb (soul) floats near her. In the background, a beam of golden light cuts through the darkness.
SCENE 3:大悲降臨
重點: 仰望天空,嘴型張大,像是在高歌祈請。
VEO 3.0 Prompt:
Low angle shot looking up at DONA. She is SINGING FERVENTLY, head tilted up towards the light, strumming the Yueqin energetically. Three magnificent pillars of golden light descend from above, illuminating her face and the white blazer. The light reaches down like welcoming hands. Divine atmosphere.
SCENE 4:脫離險境
重點: 表情放鬆,微笑著唱,眼神溫柔。
VEO 3.0 Prompt:
Medium close-up shot. DONA (18yo, blonde, AirPods Max) is SMILING GENTLY WHILE SINGING, her expression relieved and soft. She plays the Yueqin rhythmically. A protective sphere of golden light surrounds her and a glowing soul orb. Dark smoke outside the sphere dissolves upon contact with the light.
SCENE 5:一心安住
重點: 閉眼或微張眼,嘴型微動(默念感),動作非常穩定。
VEO 3.0 Prompt:
Macro shot, slow motion. DONA is sitting in a meditative pose on a giant blooming lotus, eyes half-closed, lips moving slightly as if CHANTING A MANTRA. She holds the Yueqin steadily. A pristine pink and gold lotus blooms in golden water. Absolute stillness, Zen atmosphere, golden light reflecting on her porcelain skin.
SCENE 6:光中融合
重點: 全身發光,身體隨上升氣流飄動,仍在彈奏。
VEO 3.0 Prompt:
Ascension shot. DONA, SINGING joyfully with mouth open, floats upwards on a lotus platform. She is playing the Yueqin. Her figure is bathed in intense backlight from a massive source of golden light (Amitabha) above. Her silhouette begins to merge with the infinite light. Transcendence.
SCENE 7:極樂淨土
重點: 環視四周,快樂地唱,手部彈奏動作明顯。
VEO 3.0 Prompt:
Wide panoramic shot of Pure Land with gold ground and gem trees. DONA stands in the center, SINGING HAPPILY and turning slightly to look at the falling celestial flowers. She strums the Yueqin with visible hand movements. Vibrant colors, high fantasy style, 8k resolution.
SCENE 8:持咒迴向 (循環片段)
重點: 特寫臉部,嘴型快速規律地動(配合快速唸咒),背景有咒字旋轉。
VEO 3.0 Prompt:
Close-up on DONA's face. She is CHANTING FAST AND RHYTHMICALLY, mouth moving in sync with a fast tempo. She plays the Yueqin. Glowing golden Sanskrit mantras spin gently around her in a ring of light. Background is a pulsing warm light. Hypnotic loop.
生成影片: 使用上述 8 個提示詞在 VEO 3.0 生成 8 段無聲影片。
取得音樂: 去 Suno AI 下載您生成的100秒助念歌曲 (MP3)。
剪輯 (VIDS):
開啟 VIDS 或剪輯軟體。
將 Suno 的 MP3 拉入音軌 (Audio Track)。
將 VEO 的 8 段影片拉入視訊軌 (Video Track),依序排列。
調整速度: 如果某一段影片比音樂長,使用「加速」功能;如果太短,使用「慢動作」功能,讓畫面剛好對上音樂的段落。
這樣您就會得到一部:畫面是 DONA 在極樂世界彈唱(由 VEO 生成),聲音是真實的月琴與助念(由 Suno 生成)的完美影片。