How To Get Amazing Text in Stable Diffusion - Fooocus

Подписаться 12 тыс.

Просмотров 7 тыс.

50% 1

In this video I will show you how to get great accurate text with almost any image generated with Stable Diffusion and Fooocus. I will show you the 3 different methods I use to accomplish this. You can even get small text.🎨 #Fooocus #StableDiffusion #DigitalArt #ImageManipulation #CreativeTools
Fooocus is a great interface for using Stable Diffusion and is very easy to get started. It also will run on systems with only 4GB VRAM.
I am running on an i5 32GB RAM Nvidia 3070 8GB VRAM
If you find the video useful and would like to tip, RU-vid does have the Super Thanks option. Those are greatly appreciated so I can have my coffee!!
☕ BUY ME A CUP OF ELECTRICITY: ko-fi.com/kleebztech
Patreon: / kleebztech
Fooocus Link: github.com/lllyasviel/Fooocus
How to install Fooocus video: • Stable Diffusion - Dow...
Fooocus Beginners Guide: • Super Easy Stable Diff...
Fooocus Upscales and Variations: • Stable Diffusion - Foo...
Checkpoints and LoRAs the Basics: • Stable Diffusion - Che...
Zoom with Outpainting: • Stable Diffusion - Zoo...
Fooocus Poses and FaceSwap - Image Prompts: • Stable Diffusion - Pos...
Inpainting with Fooocus - Don't Regenerate, Fix!: • Stable Diffusion - Inp...
Background Replacement with Fooocus: • Stable Diffusion - Bac...
Fooocus Tips and Tricks (and AI Hands) • Overlooked Fooocus Tip...
Stable Diffusion - Styles in Fooocus - How do they work?: • Stable Diffusion - Sty...
FaceSwap and Consistent Character TipsPart 1: • Stable Diffusion - Fac...
FaceSwap and Consistent Character Tips - Part 2: • Stable Diffusion - Fac...
00:00 Intro
00:25 Method One
06:25 Method Two
11:00 Method Three

Наука

Опубликовано:

11 июл 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 60

@buffalonan Месяц назад

There was a feature that needs to be enabled first: Mixing Image Prompt and Inpaint. It is in Debug > Control. Thanks I learned a lot!

@KLEEBZTECH Месяц назад

Thanks! Now realize I didn't mention that. Thankfully I do have it shown at a point in the video. I sometimes forget to mention the basics when doing an advanced video.

@vspolley 2 месяца назад

Here was I thinking that getting legible text was a lost cause!

@KLEEBZTECH 2 месяца назад

I have been very happy with these methods once I figured them out. That name tag was something I had not tried before doing the video. Now I need to see how small I can go.

@g_trep 2 месяца назад

I'll start sounding like a broken record, your videos are amazing. Good job 👌

@KLEEBZTECH 2 месяца назад

Thank you!

@xjohnny1000 2 месяца назад

I can't be the only one to notice that you typed "Tonya" but asked for "Tanya". Fooocus still gave the correct text anyways so that's impressive.

@KLEEBZTECH 2 месяца назад

Damn typo. 😆 I missed that even after watching it 20+ times in editing. But that does show why it is also important to put the text in the text prompt you want. Yes it corrected my typo since it was pretty close to the same shape. I am sure if it was most other letters it might not have.

@Solar546 2 месяца назад

Wow I didn't think Fooocus could do these things, thanks!

@Art0691p 2 месяца назад

Great tutorial. Clear and to the point. Hopefully in SD3 this will be easier - that seems to be the expectation. Keep up the great work!

@KLEEBZTECH 2 месяца назад

Fingers crossed!

@fishpickles1377 2 месяца назад

Wow, this is going to be SUPER helpful, what an eye opener! Thanks!

@KLEEBZTECH 2 месяца назад

Figured it would help with what you were trying to do. 👍

@richardbradley2641 2 месяца назад

Thanks! The text video was really helpful

@KLEEBZTECH 2 месяца назад

Thank you very much!

@mrbabyhugh 2 месяца назад

5:51 a stop of about 0.85 seem to work well to give as exact to image prompt as possible. weight can be 0.975-1

@xnovamix 2 месяца назад

This is awesome. Great tutorial!

@KLEEBZTECH 2 месяца назад

Thanks! It really is useful and happy I finally figured out how to do it.

@liufei3405 2 месяца назад

Master Fooocus!

@b.s.design 2 месяца назад

Great job! Thank you!

@KLEEBZTECH 2 месяца назад

Very welcome as always! Get your image grid worked out?

@b.s.design 2 месяца назад

@@KLEEBZTECH Yes. It turned out to be pretty simple ))

@b.s.design 2 месяца назад

@@KLEEBZTECH I have a counter question for you: Do you generate videos? If you do, in which generator? ForgeUI or ComfyUI? Or in some other generator?

@KLEEBZTECH 2 месяца назад

I haven't messed much with videos lately. I've been spending way too much time trying to get every ounce out of Fooocus. I had messed with them a while back and wasn't satisfied with where it was at at the time. I'll probably jump back into videos soon.

@KLEEBZTECH 2 месяца назад

If you find the video useful and would like to tip you can buy me some electricity for all these image generations. It is greatly appreciated! ko-fi.com/kleebztech

@Designing-hc5pz 2 месяца назад

thx for the tutorial i also appreciate the thumbnail 😁i've a question that how can we fix hands if it's deformed and wearing riding gloves... any idea??

@KLEEBZTECH 2 месяца назад

Yes I have ideas... You might want to check out this video ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-kpD5_Bs9Qeo.html about denoising strength and how you can paint a correction to influence inpainting. It is similar to how the improve detail function of inpainting works. If you give it a general form to work with it can be done.

@Designing-hc5pz 2 месяца назад

@@KLEEBZTECH ok thx 😀let me give it try

@KLEEBZTECH 2 месяца назад

And you could Implement part of what's in this video here for that hand.

@adithyansreeni7491 2 месяца назад

Bro is there any way to know what font is generated by the ai or a font that is similar to the generated text..actually does the font matter when we create the sample on photoshop?

@KLEEBZTECH 2 месяца назад

The font you use for guiding it with the image prompt will usually carry over more if you increase the sliders on the image prompt. Handwriting fonts are much more difficult. Bold fonts tend to work better.

@adithyansreeni7491 2 месяца назад

@@KLEEBZTECH okk vro new sub😌👍

@PawFromTheBroons 2 месяца назад

To be fair, if I'm to round-trip to Photoshop, I might as well do it there. If only to have access to the font and color of my choosing.

@KLEEBZTECH 2 месяца назад

I would definitely say if your Photoshop skills are good enough to do it quickly and well that's what I would go with. It would take me a while to get it to look good myself.

@marshallodom1388 2 месяца назад

Speaking of text... I've played with getting Text Gen Web UI to generate images with stable diffusion on the same browser, but can't seem to get it find the checkpoints. I'm about to give up, maybe in your vast array of skills you could tell me how to get Text Gen to use fooocus?

@KLEEBZTECH 2 месяца назад

I don't know if there's any way to use them together. I think that needs to use A1111.

@peterhoang2107 2 месяца назад

I can't seem to get method one working. It's either going to modify the "inpaint" or the "image prompt" BUT not both like how you done it in your video.

@KLEEBZTECH 2 месяца назад

I am not sure I understand what you mean.

@peterhoang2107 2 месяца назад

@@KLEEBZTECH I figured it out. In Debug Mode, I had enable "Mixing Image Prompt and Vary/Upscale" under the Control tab

@KLEEBZTECH 2 месяца назад

I hate when I do that. lol. @@peterhoang2107

@kromecrowell 2 месяца назад

How about vertical text?

@KLEEBZTECH 2 месяца назад

I don't see why not.

@FactQuill-Channel 2 месяца назад

Why doesn't my fooocus have a public link, and I'm told to download the file "frpc Linux amd64" What is that for? Does anyone know how to solve it?

@KLEEBZTECH 2 месяца назад

How are you running fooocus? Would need a little more information and I might be able to help.

@FactQuill-Channel 2 месяца назад

@@KLEEBZTECH I run it using a android device

@KLEEBZTECH 2 месяца назад

You're running it on an Android device? If that's the case I have no idea. You're not running it through a service? I'll admit your statement has me confused since running it on Android as far as I know is not possible.

@fredpourlesintimes 2 месяца назад

I don't have your options. Doesn't work at all

@KLEEBZTECH 2 месяца назад

Which options? Are you using Fooocus? Is it the standard Fooocus? Did you check off the advanced checkbox at the bottom of image prompt? Or the debug options?

@techmartsstore72 2 месяца назад

Amazing bro 🥵🔥🔥

@KLEEBZTECH 2 месяца назад

Thanks 🔥

@techmartsstore72 2 месяца назад

@@KLEEBZTECH i am not getting public url in google colab fooocus, can you help me, fooocus not running in local url

@tomasm1233 2 месяца назад

You could invert the text you do in photoshop just by inverting whole area, or do it white text on black background for this particular case, so when focus renders the image , text comes bright and glowy on the dark background. I know its a small thing, You already did a great job. A was struggling ading text couple days ago and here we go, your video pops up. Thanks a lot!

@KLEEBZTECH 2 месяца назад

Yeah there's so many different ways you can manipulate the text. I just recently discovered all this. I've got other things that utilize the same method. Not text related

@KLEEBZTECH 2 месяца назад

My big thing is the ability to do small text. Until recently I couldn't do it. I got some other things to try as well when it comes to text.

@tomasm1233 2 месяца назад

@@KLEEBZTECH I was wondering if training Lora solely for the inpainting text is the solution.

@tomasm1233 2 месяца назад

@@KLEEBZTECH Maybe going bigger text and then resizing it in Photoshop. I know it's a jump in jump out method, though.