Fooocus Enhance Masking Made Easy

[@MonzonMedia]

Let me know what kind of tutorials you'd like to see on Fooocus!

[@mik3lang3lo]

Are they going to add Flux support in Fooocus?

[@MonzonMedia]

Not on the original Fooocus unfortunately, it would be too much work since it's built different than SDXL. There is a fork of Fooocus that can use Flux models apparently but I haven't tried it yet.

[@Know_It_All.]

flux and video editing , creating bye ai

[@ОльгаФед-д6с]

Tell about upscaling in Fooocus. Often when you zoom in, the details and style change a lot, or the picture turns out ugly. Especially if they are drawn pictures and art.

[@pawelthe1606]

Perspective , how to manipulate how far or close object can be created in picture please

[@oskarmeszar8001]

you are the man, very nice tutorial 👌

[@Aininja94]

Nice explanation, and yes fooocus is still used by many.

[@MonzonMedia]

Absolutely!

[@GarrettXPrime]

best youtube channel in this space

[@MonzonMedia]

I appreciate that...very kind of you! 😊🙌

[@pawelthe1606]

Great video as always !!

[@MonzonMedia]

Appreciate that and thanks for the support!

[@nicdisan]

Thank you so much, great tutorial, very clear as always

[@MonzonMedia]

You’re very welcome my friend! Appreciate the support. 👍🏼🙌🏼

[@CreativeReality]

I have the latest stable diffusion webui forge, but I cannot see anything about foocus, enhance or anything related to what you´re showing. However, I do have the extension enabled

[@MonzonMedia]

Forge and Fooocus are separate platforms. To use the enhance feature you need to install Fooocus.

[@Beauty.and.FashionPhotographer]

could you inpaint and do with inpainting a face-swap , but done with a Face Lora from a person ? so not face-swapping with a single image of that face and that person, but with a face Lora from that same real person.

[@NotThatOlivia]

same thing can be done in a1111 and even better and in more granular way in ComfyUI

[@MonzonMedia]

Well sure but this is Fooocus 😊 Not everyone like to use ComfyUI or A1111

[@NotThatOlivia]

@@MonzonMedia I love all of them: a1111, forge, comfy and fooocus

[@xhridhar]

Is there any way to add a specific image of the dress to put on her rather than just prompt a color change ?

[@MonzonMedia]

To some extent yes. Working in a video for that 👍🏼

[@yunusyuruk9035]

Dostum Türkçe altyazılı da ekleseydin keşke. Söylediklerini anlayamıyorum 😢

[@MonzonMedia]

You can click on the gear icon, choose sub-titles, auto-generate and choose your language. Hope that helps!

[@StringerBell]

Foocus is no longer relevant, everybody moved long ago to Flux and the small amount of people that are still holding on SDXL will soon move on too. Too bad it's technically possibly for Fooocus to support Flux :(

[@MonzonMedia]

Not true at all my friend, There are still many people who use SD1.5 and SDXL since there is still more support, smaller file sizes, more models, lora's controlnet etc. Also not everyone has the system that can run Flux. There is a fork of Fooocus that can run Flux but I haven't tried it yet.

[@b.s.design]

Flux is a great model, but it still doesn't understand and can't do many things that SDXL models do perfectly well without any tricks. So Fooocus will be relevant and in demand for a long time to come.

[@StringerBell]

​@@b.s.design As a one of the bigger SDXL model developers on CIVITAI, I assure you that what are you saying has no connection with the reality. The prompt understanding in FLUX compared to SDXL is not even comparable. Totally different architectures too(unet vs transformers) It's like comparing a bycicle to latest BMW in terms of transportation.

[@b.s.design]

@@StringerBell You didn’t understand what I was talking about. Let me give you an example: Recently, I needed to create a video in the style of illustrator Mike Deodato. So, Flux is completely unfamiliar with this artist's style (or others), while SDXL models immediately generated his style for me without Lora’s and additional prompts. What I mean is that without additional Loras, the Flux model doesn’t understand artistic styles, nor the styles of artists, sculptors, designers, architects, etc. Undoubtedly, Flux is a major breakthrough in generative art, but the point is that while it does everything very well, unfortunately, it’s not always what you need... Now do you understand what I meant?

[@StringerBell]

@@b.s.design The base SDXL model will never be able to generate the style. You are comparing OVERTUNED models for a year to base model, released a month ago.