中文|English

Audio Processing Related ComfyUI Nodes

Audio is the bridge connecting text, video, and images. Videos without audio or text are bland. This project currently includes the following main nodes:

• Pause workflow anywhere
• Audio loading, customizable loading path, including subdirectories
  • Please rename the file extra_help_file.yaml.example to extra_help_file.yaml and remove the annotation # . Add a custom loading directory such as audios_dir: D:\AIGC\ComfyUI-Data\audios_input, Linux is /.
• String show and editing. If editing is enabled, the workflow will pause execution. After editing, you can click "Continue Workflow" to continue execution
• Automatically add subtitles to videos
• Arbitrary time scale audio cropping
• Audio volume, speed, pitch, echo processing, etc.
• Remove silent parts from audio
• Recording
• Audio Watermark Embedding

Examples:

• Pause workflow anywhere:

• Audio loading:

• String show and editing. If editing is enabled, the workflow will pause execution. After editing, you can click "Continue Workflow" to continue execution:

• Add subtitles to video:

• Combine ComfyUI_EraX-WoW-Turbo for automatic speech recognition, and then add subtitles to the video:

• Combine ComfyUI_EraX-WoW-Turbo, ComfyUI_gemmax, ComfyUI_SparkTTS, ComfyUI-LatentSyncWrapper for automatic speech recognition, automatic translation, automatic voice cloning, automatic lip sync, automatic subtitle addition to video (detailed example workflow workflow-examples):

• Arbitrary time scale cropping of audio:

• Audio volume, speed, pitch, echo processing, etc.:

• Remove silent parts from audio and recording:

• Audio Watermark Embedding (Disable watermark embedding; if a watermark exists, it will be automatically detected):

1, To use this node, download all SilentCipher models and place them in the ComfyUI\models\TTS\SilentCipher\44_1_khz\73999_iteration directory. 2, And execute pip install git+https://github.com/SesameAILabs/silentcipher install silentcipher.

📣 Updates

[2025-05-23]⚒️: Fixed the logic issue with the pause node. Now, both serial and parallel execution of pause nodes will pause for the first time. If there is no change in the preceding node, it will automatically pass again.

[2025-04-28]⚒️: Audio loading, customizable loading path, including subdirectories.

[2025-04-26]⚒️: Pause workflow anywhere.

[2025-04-25]⚒️: String show and editing.

[2025-03-28]⚒️: Added watermark embedding node.

[2025-03-26]⚒️: Released version v1.0.0.

Installation

Install sox and add it to the system path.

cd ComfyUI/custom_nodes
git clone https://github.com/billwuhao/ComfyUI_AudioTools.git
cd ComfyUI_AudioTools
pip install -r requirements.txt

# python_embeded
./python_embeded/python.exe -m pip install -r requirements.txt