Slop Demon's MMAudio
Studio
The MMAudio Studio workflow is designed to allow you to create videos with NSFW audio combined with background audio to provide a more dynamic video to audio experience.
Utilizes both MMAudio NSFW Edition and the Standard MMAudio
NSFW MMAudio nodes utilizes Image to Audio while the Standard MMAudio nodes utilize Text to Audio.
Optimizes videos to MMAudio's training to help improve audio syncing.
Downscales videos to help save VRAM and produce audio faster. (Does not impact the final video resolution)
Adds empty frames to videos that are less than 8 seconds long (Optimization to improve audio syncing, empty frames are removed before the final video output).
Provides 3 NSFW Audio Channels and 3 video outputs at once.
Provides 3 Standard MMAudio Background nodes combined with each NSFW output to provide a more dynamic audio experience. (1 NSFW Audio + 3 Background Noise Audio)
Individual Seed and Volume/Gain controls in one central location.
Individual Audio previews for each audio channel.
Color coded to provide an easy to understand connection between nodes.
NEW! Mute Toggles have been added alongside an Output Muter option to bypass the video outputs.
Solo
Solo edition is a more compact version of studio offering a single output rather than three for fast single outputs. This is the newest version and now includes two new options.
Mute Toggles have been added that allow you to mute a single audio output without needing to alter your current volume.
This edition has also been setup to utilize ComfyUI's new App Mode option. The workflow will still start in the default nodes setup, but you can enter app mode if you prefer that interface.
Zero
Zero is the latest workflow taking the Solo version as a base this edition compacts the entire workflow so all of your controls are within much closer proximity allowing you to make changes quickly and efficiently.
This version comes with a minor improvement now the Mute toggle will no longer mute the audio preview and will only mute the audio for the final output.
Updates
Zero v2 released 3/24/26
Zero v2 offers some improvements to the original
Toggles have been added to each Prompt allowing you to choose between Video to Audio or Text to Audio and the ability to toggle between the NSFW mmaudio model or the mmaudio Standard model
Gain dials have been updated to 1 step increments instead of 5 allowing you to finely tune the volume
Labels have been updated and the Load and Preprocess section has been condensed and reorganized.
Both Studio and Solo have been updated 3/13/26
Studio Update
Studio has been reorganized, received Mute toggles, and Output Muter's.
Solo Update
Mute toggles have been fixed, image preprocessing is now dependent on the video's width and height and will be downscaled by 1/2 by default to save VRAM and speed up processing.
Models
MMAudio Standard
https://github.com/hkchengrex/MMAudio
MMAudio NSFW
https://huggingface.co/cloud19/NSFW_MMaudio/tree/main?not-for-all-audiences=true
Description
FAQ
Comments (3)
From what I see in your samples, this is supposed to work something like: channel 1: woman moaning, channel 2: street ambiance sound, channel 3 music... something like that?
I mention it in the description on the main page. Channel 1 uses the NSFW edition of MMAudio and uses your video input. The 3 additional channels use the standard MMAudio and do not rely on your video input to produce background ambiance whether that's music or sound effects is up to you.
@SlopDemon Ok, clearer now, thanks!

