Invoke AI 5.0 Tutorial - From Beginner to Pro in Minutes! Part 2 Let's Get Creative!

[@MonzonMedia]

If you are new to Invoke AI be sure to check out Part 1 Navigation & Settings kzbin.info/www/bejne/jWXRg4BnnqZsiNksi=KdP1lKgbODhmrAR2

[@satoricanton]

Very good! I really needed to know how to move around the canvas. I knew I was missing something. Didn't know to hit the space bar to move around. Thanks so much! Can't wait to see more

[@MonzonMedia]

@@satoricanton the hot keys are similar to PG canvas. My next vid will be on using regional guidance and ControlNet! 👊🏼🙌🏼

[@captainpike3490]

Invoke's canvas features, especially with its control layers, are designed to be highly beneficial for illustrators and graphic designers, offering a familiar and efficient workflow similar to design softwares. Canvas features deserve series of tutorials there is a lot to learn, you can go very creative. Thank you

[@MonzonMedia]

Agreed! It's very powerful and I've always felt that it was the best ai image editing tool currently out there. Even better that it's open source!

[@tomschuelke7955]

its not only great for designer... its also extremly beneficial for architectural workflows. here, in contrast to Art and Design , you have to follow design ideas much more strictly.. its extremly difficula to reproduce a panorama view of a realy big architectural composition... i mean.. smal rooms, smal houses are not a problem. But when you are in Architectural international competitions, i think only Invoke AI even starts... to deliver the control you have.. Now you can import not only 1024 pix images, but also 4K CAD Screenshots.. , use them als controlnets, force invoke by "manualy" downscaling the image to the perfect sice of the ai.. and afterwords detailing the areas of the image in 4K, becaus your controlnet in 4K is still available. This feature alone for real and big architecture is worth the upgrade to Invoke 5. We where able to improve the detail quality of our ai images very much this way.. Of course you can still in the very early design phase of the architecture pic up great and random idas the ai produces by accident, just by loosing your ip adapters and controlnets.. And with flux now also available ( at the moment still without controlnets) you can improve parts even furthere with img to img or regional inpainting an SDXL result..

[@guidogroen2536]

Yes! Very well made. Much appreciated!

[@MonzonMedia]

Much appreciated!

[@tomschuelke7955]

great... one thing i think now works.. when you save your canvas image right top of the canvas, the saved image can recreate all the layers in the canvas.. But! only as long as you dont delete the in between images that are allways storred invisible in the database too.. so its a bit more possible evevn if you clear your canvas some time

[@MonzonMedia]

Ooooooh nice! Will try it out. I knew there had to be a way! 😬👊🙌

[@emv2280]

Hey Monzon! As always you're my go-to for everything AI! Would you say this is sort of similar to how Playground AI used to be before the redesign? I've been looking for a new "more advanced on the user side" AI to swap too after the latest changes, that has all of their previous canvas features, but that isn't too cartoony with their filters. I like to create real models for reference. Any tips?

[@MonzonMedia]

Hey there! While it doesn't have some features from PG like object eraser or live preview it is a more advanced Canvas. You can use any model from civitai.com. You can find a lot of the ones that were on PG but for realism RealvisXL, Realism Engine, Epic Realism XL are all good for that. Juggernaut XL is a good all around one.

[@firemageddon]

Awesome vid. Thank you.

[@MonzonMedia]

You’re welcome! More to come soon!

[@streetphone4619]

Thanks. What a great app!

[@MonzonMedia]

@@streetphone4619 Welcome! It’s a very powerful and underrated tool.

[@cyberbol]

I trained my model, checkpoint using dreambox flux. I converted from 23gb to 11gb. It works fine on SwarmUI, on FORGE as well, but not on INVOKE. I can install it, but when i try generate anything invoke just restart it self in console.

[@MonzonMedia]

Are you using the text encoders that invoke ai provides? As it stands you have to use theirs, should be in the starter models section. Download it and load it in the advanced section.

[@aniketgore0]

Nice video

[@MonzonMedia]

Thank you!

[@gar-r3w3b]

nice job thank

[@MonzonMedia]

You're welcome! Thanks for checking it out! 🙌

[@Maylin-ze6qx]

❤❤❤❤

[@MonzonMedia]

🙌😊

[@giovanni.dipinto]

does it have a faceswap like the one on Fooocus ?

[@MonzonMedia]

It can use IP adapter for face which is similar. Will show that soon! 👍

[@tomschuelke7955]

you can use ip adapters... yes. but invoke does strictly restrict software to only free software... so the best ones available are not free. and thats why they are not yet integrated. Invoke fokuses strongly on total legal use... which is fine when you work professionaly..

[@giovanni.dipinto]

@@tomschuelke7955 i don't know well Invoke ai for now, does it have image2image ?

[@giovanni.dipinto]

or better, image prompt as Foocus ?