

Powered By RTX 4090 Series
LU_短剧六件套之二:图生分镜(GPT-4提示词/ div提示词)
生成多张分镜图片, 用于图生视频, 多帧生视频的前置,
如果ai生成提示词只需要输入主题和多少张分镜, 选择true:ai prompt.
如果自己填提示词只需要输入定义提示词, 选择false:div prompt, 自己输入提示词, 一行代表一张.生成速度取决于分辨率, 一般1K的9张耗时是: 3分58秒
ps: 如果使用gpt 4生成提示词就只需要填主题和需要多少分镜. 如果要自定义提示词, 就自己填, 一行一张分镜图.
ps:一般我是执行多次(5-10次吧), 生成大概45-90张图片, 然后挑选大概20张左右当做分镜头的起始帧.
------------------------------------------------------------------------
LU_Two of the six pieces of the short drama set: image-to-storyboard(GPT-4 prompt/ div prompt.)
Generate multiple storyboard images for use in image generated videos,
If AI only needs to input the theme and how many storyboards to generate prompt words, choose 'true: AI prompt'
If you fill in the prompt yourself, just enter the definition prompt and select 'false: div prompt'.Enter the prompt words yourself, with each line representing one image The generation speed depends on the resolution, and the typical time for 9 frames of 1K is: 3min 58s
PS: If using GPT 4 to generate prompt words, you only need to fill in the theme and how many storyboards are needed If you want to customize the prompt words, just fill them in by yourself, with one storyboard image per line
PS: Usually, I execute it multiple times (5-10 times), generate about 45-90 images, and then select about 20 of them as the starting frames for the storyboard
------------------------------------------------------------------------
H-version等之后会出.
LU_短剧六件套之一:文生图
https://www.runninghub.cn/post/1987052252043087873
LU_短剧六件套之二:图生分镜
https://www.runninghub.cn/post/1988519232235331585
LU_短剧六件套之三:图生视频
https://www.runninghub.cn/post/1987043435041460226
LU_短剧六件套之四:首尾帧生视频
https://www.runninghub.cn/post/1987043353088954370
LU_短剧六件套之五:多帧生视频
https://www.runninghub.cn/post/1987043344478089218
LU_短剧六件套之六:视频合成+音频合成.
https://www.runninghub.cn/post/1987052320276066306
------------------------------------------------------------------------
H-version, It will be released later.
LU_One of the six pieces of the short drama set: text-to-image
https://www.runninghub.ai/post/1987052252043087873
LU_Two of the six pieces of the short drama set: image-to-storyboard
https://www.runninghub.ai/post/1988519232235331585
LU_Three of the six pieces of the short drama set: image-to-video
https://www.runninghub.ai/post/1987043435041460226
LU_Four of the six pieces of the short drama set: first-Last-to-video
https://www.runninghub.ai/post/1987043353088954370
LU_Five of the six pieces of the short drama set: multi-frame-to-video
https://www.runninghub.ai/post/1987043344478089218
LU_Six of the six pieces of the short drama set: Video Audio Synthesis
https://www.runninghub.ai/post/1987052320276066306
生成多张分镜图片, 用于图生视频, 多帧生视频的前置,
如果ai生成提示词只需要输入主题和多少张分镜, 选择true:ai prompt.
如果自己填提示词只需要输入定义提示词, 选择false:div prompt, 自己输入提示词, 一行代表一张.生成速度取决于分辨率, 一般1K的9张耗时是: 3分58秒
ps: 如果使用gpt 4生成提示词就只需要填主题和需要多少分镜. 如果要自定义提示词, 就自己填, 一行一张分镜图.
ps:一般我是执行多次(5-10次吧), 生成大概45-90张图片, 然后挑选大概20张左右当做分镜头的起始帧.
------------------------------------------------------------------------
LU_Two of the six pieces of the short drama set: image-to-storyboard(GPT-4 prompt/ div prompt.)
Generate multiple storyboard images for use in image generated videos,
If AI only needs to input the theme and how many storyboards to generate prompt words, choose 'true: AI prompt'
If you fill in the prompt yourself, just enter the definition prompt and select 'false: div prompt'.Enter the prompt words yourself, with each line representing one image The generation speed depends on the resolution, and the typical time for 9 frames of 1K is: 3min 58s
PS: If using GPT 4 to generate prompt words, you only need to fill in the theme and how many storyboards are needed If you want to customize the prompt words, just fill them in by yourself, with one storyboard image per line
PS: Usually, I execute it multiple times (5-10 times), generate about 45-90 images, and then select about 20 of them as the starting frames for the storyboard
------------------------------------------------------------------------
H-version等之后会出.
LU_短剧六件套之一:文生图
https://www.runninghub.cn/post/1987052252043087873
LU_短剧六件套之二:图生分镜
https://www.runninghub.cn/post/1988519232235331585
LU_短剧六件套之三:图生视频
https://www.runninghub.cn/post/1987043435041460226
LU_短剧六件套之四:首尾帧生视频
https://www.runninghub.cn/post/1987043353088954370
LU_短剧六件套之五:多帧生视频
https://www.runninghub.cn/post/1987043344478089218
LU_短剧六件套之六:视频合成+音频合成.
https://www.runninghub.cn/post/1987052320276066306
------------------------------------------------------------------------
H-version, It will be released later.
LU_One of the six pieces of the short drama set: text-to-image
https://www.runninghub.ai/post/1987052252043087873
LU_Two of the six pieces of the short drama set: image-to-storyboard
https://www.runninghub.ai/post/1988519232235331585
LU_Three of the six pieces of the short drama set: image-to-video
https://www.runninghub.ai/post/1987043435041460226
LU_Four of the six pieces of the short drama set: first-Last-to-video
https://www.runninghub.ai/post/1987043353088954370
LU_Five of the six pieces of the short drama set: multi-frame-to-video
https://www.runninghub.ai/post/1987043344478089218
LU_Six of the six pieces of the short drama set: Video Audio Synthesis
https://www.runninghub.ai/post/1987052320276066306
lu
2025-11-15 更新
lu
2025-11-15 更新
工作流介绍
LU_短剧六件套之二:图生分镜(GPT-4提示词/ div提示词)
生成多张分镜图片, 用于图生视频, 多帧生视频的前置,
如果ai生成提示词只需要输入主题和多少张分镜, 选择true:ai prompt.
如果自己填提示词只需要输入定义提示词, 选择false:div prompt, 自己输入提示词, 一行代表一张.生成速度取决于分辨率, 一般1K的9张耗时是: 3分58秒
ps: 如果使用gpt 4生成提示词就只需要填主题和需要多少分镜. 如果要自定义提示词, 就自己填, 一行一张分镜图.
ps:一般我是执行多次(5-10次吧), 生成大概45-90张图片, 然后挑选大概20张左右当做分镜头的起始帧.
------------------------------------------------------------------------
LU_Two of the six pieces of the short drama set: image-to-storyboard(GPT-4 prompt/ div prompt.)
Generate multiple storyboard images for use in image generated videos,
If AI only needs to input the theme and how many storyboards to generate prompt words, choose 'true: AI prompt'
If you fill in the prompt yourself, just enter the definition prompt and select 'false: div prompt'.Enter the prompt words yourself, with each line representing one image The generation speed depends on the resolution, and the typical time for 9 frames of 1K is: 3min 58s
PS: If using GPT 4 to generate prompt words, you only need to fill in the theme and how many storyboards are needed If you want to customize the prompt words, just fill them in by yourself, with one storyboard image per line
PS: Usually, I execute it multiple times (5-10 times), generate about 45-90 images, and then select about 20 of them as the starting frames for the storyboard
------------------------------------------------------------------------
H-version等之后会出.
LU_短剧六件套之一:文生图
https://www.runninghub.cn/post/1987052252043087873
LU_短剧六件套之二:图生分镜
https://www.runninghub.cn/post/1988519232235331585
LU_短剧六件套之三:图生视频
https://www.runninghub.cn/post/1987043435041460226
LU_短剧六件套之四:首尾帧生视频
https://www.runninghub.cn/post/1987043353088954370
LU_短剧六件套之五:多帧生视频
https://www.runninghub.cn/post/1987043344478089218
LU_短剧六件套之六:视频合成+音频合成.
https://www.runninghub.cn/post/1987052320276066306
------------------------------------------------------------------------
H-version, It will be released later.
LU_One of the six pieces of the short drama set: text-to-image
https://www.runninghub.ai/post/1987052252043087873
LU_Two of the six pieces of the short drama set: image-to-storyboard
https://www.runninghub.ai/post/1988519232235331585
LU_Three of the six pieces of the short drama set: image-to-video
https://www.runninghub.ai/post/1987043435041460226
LU_Four of the six pieces of the short drama set: first-Last-to-video
https://www.runninghub.ai/post/1987043353088954370
LU_Five of the six pieces of the short drama set: multi-frame-to-video
https://www.runninghub.ai/post/1987043344478089218
LU_Six of the six pieces of the short drama set: Video Audio Synthesis
https://www.runninghub.ai/post/1987052320276066306
生成多张分镜图片, 用于图生视频, 多帧生视频的前置,
如果ai生成提示词只需要输入主题和多少张分镜, 选择true:ai prompt.
如果自己填提示词只需要输入定义提示词, 选择false:div prompt, 自己输入提示词, 一行代表一张.生成速度取决于分辨率, 一般1K的9张耗时是: 3分58秒
ps: 如果使用gpt 4生成提示词就只需要填主题和需要多少分镜. 如果要自定义提示词, 就自己填, 一行一张分镜图.
ps:一般我是执行多次(5-10次吧), 生成大概45-90张图片, 然后挑选大概20张左右当做分镜头的起始帧.
------------------------------------------------------------------------
LU_Two of the six pieces of the short drama set: image-to-storyboard(GPT-4 prompt/ div prompt.)
Generate multiple storyboard images for use in image generated videos,
If AI only needs to input the theme and how many storyboards to generate prompt words, choose 'true: AI prompt'
If you fill in the prompt yourself, just enter the definition prompt and select 'false: div prompt'.Enter the prompt words yourself, with each line representing one image The generation speed depends on the resolution, and the typical time for 9 frames of 1K is: 3min 58s
PS: If using GPT 4 to generate prompt words, you only need to fill in the theme and how many storyboards are needed If you want to customize the prompt words, just fill them in by yourself, with one storyboard image per line
PS: Usually, I execute it multiple times (5-10 times), generate about 45-90 images, and then select about 20 of them as the starting frames for the storyboard
------------------------------------------------------------------------
H-version等之后会出.
LU_短剧六件套之一:文生图
https://www.runninghub.cn/post/1987052252043087873
LU_短剧六件套之二:图生分镜
https://www.runninghub.cn/post/1988519232235331585
LU_短剧六件套之三:图生视频
https://www.runninghub.cn/post/1987043435041460226
LU_短剧六件套之四:首尾帧生视频
https://www.runninghub.cn/post/1987043353088954370
LU_短剧六件套之五:多帧生视频
https://www.runninghub.cn/post/1987043344478089218
LU_短剧六件套之六:视频合成+音频合成.
https://www.runninghub.cn/post/1987052320276066306
------------------------------------------------------------------------
H-version, It will be released later.
LU_One of the six pieces of the short drama set: text-to-image
https://www.runninghub.ai/post/1987052252043087873
LU_Two of the six pieces of the short drama set: image-to-storyboard
https://www.runninghub.ai/post/1988519232235331585
LU_Three of the six pieces of the short drama set: image-to-video
https://www.runninghub.ai/post/1987043435041460226
LU_Four of the six pieces of the short drama set: first-Last-to-video
https://www.runninghub.ai/post/1987043353088954370
LU_Five of the six pieces of the short drama set: multi-frame-to-video
https://www.runninghub.ai/post/1987043344478089218
LU_Six of the six pieces of the short drama set: Video Audio Synthesis
https://www.runninghub.ai/post/1987052320276066306
节点信息
22
CLIPLoader
KSampler
LoadImage
LoraLoaderModelOnly
SaveImage
UNETLoader
VAEDecode
VAELoader
CFGNorm
Int
JDCN_StringToList
JjkText
LayerUtility: ImageScaleByAspectRatio V2
ModelSamplingAuraFlow
PrimitiveBoolean
ProcessString
RH_LLMAPI_NODE
Text Concatenate
TextEncodeQwenImageEdit
TextEncodeQwenImageEditPlusAdvance_lrzjason
easy ifElse
easy showAnything