Add sync audio to video generation with MMAudio: Installation and workflow included.

[@fizababar4508]

you elaborate this whole in a very simple way. I explored a lot of other youtubers videos for comfyUI but the basic details they always missed or if they tell but in a tricky way that for beginner who is new in comfyUI like me always messed up the things. Thanks alot for this great value.

[@CodeCraftersCorner]

Thank You! Glad you found it helpful!

[@MrRandomPlays_1987]

Thank you, Im a noob with this whole ComfyUI stuff so it's exactly the video I was looking for and after a quick search in YT I found your video, thanks.

[@CodeCraftersCorner]

Awesome! I'm glad you found it helpful.

[@CalladsEssence]

Hi Sharvin! Works like dream! Thanks again for your tutorial!

[@CodeCraftersCorner]

Thank you! Glad it helped!

[@AINMEisONE]

Thank you! This is so amazing! I am checking your workflow!

[@CodeCraftersCorner]

Thank you. Be sure to have the correct Torch version install for it to work.

[@CamiloAguilar-nr6wj]

Thank you!!!!!!

[@CodeCraftersCorner]

Glad you found it helpful!

[@Winale]

I have a ComfyUI installation with AMD and miniconda When I launch comfyUI it tells me an error: ModuleNotFoundError: No module named 'torchdiffeq' Cannot import C:\Software\ComfyUI\custom_nodes\ComfyUI-MMAudio module for custom nodes: No module named 'torchdiffeq' But actually, torchdiffec is installed in the miniconda folder This is very specific, but maybe you have an idea what to do? Thank you so much!

[@CodeCraftersCorner]

Since you have an AMD card, it's possible the version required is different. the official GitHub says that Torch version should be 2.5.1. You may have to upgrade yours. It might be worth a try. Sorry, not sure since I cannot test without an AMD card.

[@aivideos322]

Go into comfy ui dir, update folder, use the update comfy and python dependencies.

[@CodeCraftersCorner]

@@aivideos322 It does not seem like @Winale have the ComfyUI portable version from the folder path and the mentioned of miniconda. Updating the torch version manually from the terminal might help.

[@sudabadri7051]

Thanks very nice

[@CodeCraftersCorner]

Thank you! Happy to hear!

[@Ashabarala]

Hello and thank you very much for this informative video. I have ComfyUI on Stability Matrix. Could you please tell me, How do I install the dependencies?

[@CodeCraftersCorner]

Hello, I do not have Stability Matrix. The steps will be same since it is still ComfyUI. You will have to replace the python_embeded/python.exe part with stability matrix python executable file. I am not sure where that it since I cannot test it.

[@mohamedmed4639]

When uploading the video, is there a watermark? And my second question is, are these videos available for commercial use on KZbin? Thank you.

[@CodeCraftersCorner]

No watermark and yes for commercial use. The project is under MIT.

[@mohamedmed4639]

@CodeCraftersCorner What do you mean by MIT?

[@CodeCraftersCorner]

@@mohamedmed4639 I meant the MIT License. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

[@tridhe]

Do I need to install any custom nodes other than the video helper suite? Because I am getting missing nodes when I load the workflow. I have already cloned and installed form the github repo. What could be the issue? Error: Missing Node Types When loading the graph, the following node types were not found MMAudioSampler MMAudioFeatureUtilsLoader MMAudioModelLoader

[@CodeCraftersCorner]

Hello, you will need to install the mmaudio custom node. In the video, I showed how to install it manually since it was not available in the ComfyUI Manager. Check if you have done all the steps at 7:31 (Chapter Installation) and that there are no errors in the terminal when you restart ComfyUI.

[@kxdproductions2375]

it didn't automatically install nvidia folder in the mmaudio folder you created after opening up the workflow, says the nodes are still missing on my comfyui

[@CodeCraftersCorner]

Did you install the MMAudio Custom Nodes, properly? Like I said, it required torch 2.5.1 which is present in the latest ComfyUI installation. The missing nodes are because you don't have the mmaudio install properly. I explained how to do this at 07:31

[@mohamedmed4639]

I am not skilled in using computers. Can you guide me how to install comfy ui for the initial procedure and then complete the rest of the steps to install mmaudio?

[@CodeCraftersCorner]

There is a video title: Installing ComfyUI & ComfyUI Manager made easy! Here is the link: kzbin.info/www/bejne/p3yyZ6OQdttjoK8

[@eoprodn-palmier2944]

Hi, can you help me ? i have this issue : MMAudioSampler mat1 and mat2 must have the same dtype, but got Half and Float

[@CodeCraftersCorner]

Can you check if you are using the fp16 or fp32 version. The workflow included is set up for fp16 version.

[@mohamedmed4639]

Do I have to use a specific version of Windows to download the file?

[@CodeCraftersCorner]

You will need the latest version of ComfyUI. Since you are on Windows, you can download the latest portable version. It has the necessary dependencies for this project to work.

[@mohamedmed4639]

Is there a specific number of videos I can produce daily with MMoudio?

[@CodeCraftersCorner]

No, you can generate all day, locally.

[@mohamedmed4639]

Can I download the file from a Windows 7 device?

[@CodeCraftersCorner]

You should be able to. As long as you have a GPU for the audio generation, you should be good. You will need ComfyUI to run it.

[@bestof467]

Its pronounced 🔊"AW-DIO"😂

[@pwknai]

Do you think this generator can make a sound effect for KumgFu action fight?

[@CodeCraftersCorner]

I tried it, it does not generate the sound, "yeeehh, haaahhh" stuff but maybe with some better prompting. I got it to do a broken window effect (sort of). I think it is best if you can isolate the action to a single clip and generate the sound effect, one at a time.

[@pwknai]

@@CodeCraftersCorner just in my guess... I think the generator doesn't recognize or imagine human' voice sound not like natural sound wind.. dropping... crashing... etc. either I didn't expect that could make Bruce Lee shouting hahaha.