this video and that style list is everything i've been looking for to develop a more profound applicable knowledge thanks bromigo! i'll look for your stance on flux ultra!
Thank you so much! I'm so glad this is helpful for you, and please feel free to share with others that you may find it could be viable. Much appreciated
I started a post processing workflow yesterday specifically for Flux, definitely nabbing a few of the ideas here :) For Inpaint it was based on the first Inpaint workflow I saw for comfy that just used a set Latent Noise Mask node and both of your implementations are better. Going to see if I can get differential diffusion to work with the first one and going to revisit the Fooocus Inpaint nodes as they might work with that setup. Thank you, awesome video.
That's great and I've already started tweaking as well rebuilding my normal loader workflow with elements of using demon flux model (4 steps a piece is quick for great results to then composite before doing final renderings)
@@GrocksterRox, watched FiveBelowFiveUK's Inpaint video and they're using Florence and SAM2 to create the mask, so connected those up to the Mask to SEGS node rather than doing the mask manually and it works well :)
Yes, this video opened my eyes to the greatness of flux, but I still think it's still a pretty high bar for newbies, and I wanted to ask if anyone has run FLUX on Mimicpc? How was the usage? I've been using Mimicpc for a while now, and I'd like to find some good buddies to share and learn with!
I haven't tried Mimicpc (assuming a service similar to Runpod, etc. etc.). It can be a high bar to learn, but I recommend to everyone to start with the basic foundation elements and see how they plug into each other and build on that foundation (my first videos in the playlist are a great starting point)
Thaks for this tutorial, but i have a problem loading the "flux guidance" node, the manager doesnt seem to know it exists, even after I update comfyui and try to install missing nodes, does anyone have a solution?
That's interesting - It is definitely a foundational node, so if you've updated comfy itself recently (instead of just specific nodes), you should be all set. Also just to confirm you're using standard comfy versus the upcoming/new comfyUI interface?
Thanks a bunch, but for bigginer it's always confusing to edit the workflow to get a certain result, so maybe having 1 workflow entirely for 1 option and ready to queue.
Absolutely! That's a great thought and that technique is definitely recommended for those that are new (in fact the first video I'd recommend to learn the basic building blocks is here - ruclips.net/video/NaP_PfR7qiU/видео.html)
Thanks - this is actually more for example sake to talk through particular skills, I'll be releasing a new loader with my next video which is nice and tidy and "Flux Fantastic" :)
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
this video and that style list is everything i've been looking for to develop a more profound applicable knowledge thanks bromigo! i'll look for your stance on flux ultra!
Thank you so much! I'm so glad this is helpful for you, and please feel free to share with others that you may find it could be viable. Much appreciated
Wow nice workflow. I'll have to checkout your layering and mask videos.
Definitely, a lot of great new stuff coming out very soon as well!
Do you have videos on how to do the layering?
Absolutely - ruclips.net/video/9bwhoiYR920/видео.html
@GrocksterRox thank you. I just found your channel and I'm new to comfyui. You have so much great stuff. Thanks for sharing
@@AberrantArt definitely enjoy!
This video helps a lot. Thanks!
You're welcome! Anything I can help the community with is valuable, and if you have other suggestions, feel free to let me know.
Amazing! Thank you this is so helpful!
I'm so glad this was valuable to you - feel free to share with others as the more we educate, the more great content gets produced.
I started a post processing workflow yesterday specifically for Flux, definitely nabbing a few of the ideas here :) For Inpaint it was based on the first Inpaint workflow I saw for comfy that just used a set Latent Noise Mask node and both of your implementations are better. Going to see if I can get differential diffusion to work with the first one and going to revisit the Fooocus Inpaint nodes as they might work with that setup.
Thank you, awesome video.
That's great and I've already started tweaking as well rebuilding my normal loader workflow with elements of using demon flux model (4 steps a piece is quick for great results to then composite before doing final renderings)
@@GrocksterRox, watched FiveBelowFiveUK's Inpaint video and they're using Florence and SAM2 to create the mask, so connected those up to the Mask to SEGS node rather than doing the mask manually and it works well :)
Love it!
Wow, an amazing video. Thank you! Could you give a link to the Flux Style Test Gallery? Can't find it online
Thank you so much - it's in the link on civ from this YT description but I've pasted here as well: enragedantelope.github.io/Styles-FluxDev/
@@GrocksterRox Oh sorry, didn't know that the civit link was for that as well. Thanks!
@@AI_Creatives_Toolbox all good, enjoy!
Really good video again 🎉 helping really to learn new stuff with comfy UI
Thanks again so much freek! Your models are also top notch and ahead of the curve!
Yes, this video opened my eyes to the greatness of flux, but I still think it's still a pretty high bar for newbies, and I wanted to ask if anyone has run FLUX on Mimicpc? How was the usage? I've been using Mimicpc for a while now, and I'd like to find some good buddies to share and learn with!
I haven't tried Mimicpc (assuming a service similar to Runpod, etc. etc.). It can be a high bar to learn, but I recommend to everyone to start with the basic foundation elements and see how they plug into each other and build on that foundation (my first videos in the playlist are a great starting point)
@@Huang-uj9rt A way for newbies I recommend is online generation on tensor for example made an ai tool called flux+ enhancer
Hey man! Awesome video as always.
I can't find the style list link.
Hi there it was in the link for Civ listed in the description, but here it is as well: enragedantelope.github.io/Styles-FluxDev/
Thank you for teaching us 🙏
My pleasure, if I help people, it helps to advance everyone!
Awesome! I loved the avatar at the start do you have a tutorial on this? are you using D-ID?
Thanks so much. I vary between Hedra and Live Portrait but always looking for new and easy options :)
Hello my friend,
What software do you use to create mouth gestures that match your voice? Please let me know thanks
I vary between Hedra, live portrait and a couple other stabilizer methods
You can definatelly split this videos to seperate ones
Thanks - this is good feedback!
Hey, awesome video bro! Is the avatar on the lower right made with LivePortrait or Hedra?
I flip back and forth, so this time Hedra
Works also quite well in swarmUI.lot easier for beginners or ppl not familiar with comfy nodes.
Super cool - yeah I heard swarm is good for multi-GPU operations as well
Thaks for this tutorial, but i have a problem loading the "flux guidance" node, the manager doesnt seem to know it exists, even after I update comfyui and try to install missing nodes, does anyone have a solution?
That's interesting - It is definitely a foundational node, so if you've updated comfy itself recently (instead of just specific nodes), you should be all set. Also just to confirm you're using standard comfy versus the upcoming/new comfyUI interface?
@@GrocksterRox yes, i believe this is the standard version
You can try it without the flux guidance node?
i somehow get a permission denied error when it comes to the mask part of the workflow never had smth like this before with comfy
Hmm I haven't seen that one before. Happy to take a look if you wanted to jump on the discord support channel? discord.gg/HP7xuBnPAS
even i installed MaraScott Nodes, the corresponding nodes are still missing....
If you want to jump on the discord, I might be able to quickly help diagnose your issue
good
I'll take it - always open with feedback to learn how I can turn good into great.
👋
Love it :) Thx
Thanks a bunch, but for bigginer it's always confusing to edit the workflow to get a certain result, so maybe having 1 workflow entirely for 1 option and ready to queue.
Absolutely! That's a great thought and that technique is definitely recommended for those that are new (in fact the first video I'd recommend to learn the basic building blocks is here - ruclips.net/video/NaP_PfR7qiU/видео.html)
The question is> What tool did the job for you to make the picture talk that naturally at the beginning of your video ?
Hedra
Thanks for the assist :)
👍
Appreciated
Wow ! Just looked at your Workflow. You have WAY TOO MUCH stuff in this thing...
Thanks - this is actually more for example sake to talk through particular skills, I'll be releasing a new loader with my next video which is nice and tidy and "Flux Fantastic" :)