Thanks Sebastian. I appreciate your calm style of making how-to videos, speaking clearly and explaining things to beginners, unlike other UA-cam teachers. I think other non-native English speakers will also appreciate this. It’s great that you provide good settings and show things thoroughly. I was looking for an answer on how to use the .csv style you made in Photoshop-plug-in and I got the answer. Which do you prefer to use: Photoshop use or a local web browser? Please, make a video where you demo the good/bad sides of both and give good workflow instructions.
Thank god it's finally in photoshop. It's perfect for me so I can get down to editing and using in my Visual Effects pipelines. I absolutely needed this. Thanks!
Will the prompt accept Textual Inversion keywords and Lora inside PS as well? This looks very useful, and basically what I think Adobe should have done to begin with instead of adobe Firefly
@@sebastiankamph If you are talking about the line around the edge or the file, all the pixels are filled up when I zoom in. So it will not affect printing. Or if you mean the seam between the edge of the square. For example, I change the canvas from 1024 X1024 to 1024 x 1920. When I select the empty areas, I also select a little off the edge of the square. I don't see the seem at all. Only the edges of the entire canvas, which is an optical illusion. I asked because I find that Ai generate actually picks up the style and make something similar much faster than SD. night and day difference. It took me 15min to add details to an artwork while it took me an hour to inpaint and generate each graphic. It took me hours and still couldn't achieve what I did in 15min using PS. Over 20 yrs on PS and 3 weeks on SD. I may be miss something though.
I have been using generative fill in photoshop for a little while and I've noticed it does a really good job of matching the style but, it's often in lower quality than the original. I can't get the generation as crisp as the original image. Maybe I'm missing something?
Many many thanks for the clear Installation instruction haha... Except when I setup my canvas and select 512 X 512 and hit generate. A dialog box pop up and said 'The command "show" is not currently available. Hahaha... So close yet so far. Now I don't even know how to fix? Have you got any brilliant idea?
I got it to work with stability matrix + forge but inpainting just isn't as good as it is within A1111 itself. Also my Auto-Photoshop-SD looks different than in this video. Is there an updated tutorial or is this plug in just not very good? I am also wondering if anyone has found the solution to inpainting that lets you erase within A1111? Is there an extension or something for this?
what is the "prompt shortcut" checkbox for? there is a severe lack of documentation..thanks for your great tutorials! this is a very good alternative to Invoke
I guess that would tell the plugin to use 'prompt shortcuts' or not (I've always left it on so I guess that's it?) For those wondering, prompt shortcut in this plugin is the same thing as the 'Style' dropdown menu in the Automatic1111 UI where you can save and quickly apply prompt presets. This is found in the Settings tab IIRC.
Is this for Photoshop version 2023 only (or maybe version 2022 is ok, we just need the correct plugin)? And when you said stable diffusion installation, do you mean Automatic1111? Thanks.
The VRAM is not the main concern. Photoshop can simply crash when it will eat up all remaining RAM. Automatic1111 will keep running tho. I manage to use this on 4GB VRAM with more or less pain.
@@ZKatt as you might know now, you only get a certain amount of generative tokens. Each time you press generate you use one token. If you have all apps on your subscription you get 1000 tokens and from there it’s 500, 250, 100 etc. depending. If you run out you can buy more tokens.
Krita and I believe Gimp also have something like this, but when I last looked at the repo for them, the projects looked dead and hardly get updated which is sad.
Current v1.2.3 version does not seem to be working. It connects to SD and gets LORAs list etc, but image generation just posts errors - Adobe Photoshop "The command Show is not currently available" and Adobe Photoshop "The command Set is not currently available". Is there any way to get the earlier 1.2.2 version? Other comments have been that this one still works.
@@sebastiankamph Thank you, Sebastian! I finally was able to install Stable Diffusion so will go through your other tutorials. And hopefully can connect it through API.
When I try to generate an image, I'm getting the warning popup "the SHOW command is not available" ??? Anyone have the same issue or know what the issue is and how to address it?
Plugin can't communicate with automatic1111, you forgot to add --api flag to the webui commands flags? i added it but it does't work. Why make a video without DETAILED instructions UGHHH
what a sad world we live in. i used to like digital art but i guess thats an end here. i guess the only art that still exists is the one drawn on paper with actual color you can see dry. imagine a world were ever time you see a picture or text or youtube video or whatever you need to ask if its made by a human first...sry but i m not a robot and humans letting robots create something is no art
i.imgur.com/Ah4AhEm.png here is my answer in screenshot, because youtube makes all my answer crash for some reason, maybe a word is censored here, I've no ideas what's happening...
Honestly this plug-in is very very buggy, I've never ever gotten out painting to work and in painting is hit or miss at best. I've seen the author of the plug-in do some pretty amazing things with it but he's not offering any explanations for how he does these things and there appears to be almost zero documentation
I also found it extremely challenging to find any info on this. The outpainting works ok except I haven't found a way around that weird masking issue. The inpainting is pretty good as long as you keep the selection square.
Thanks Sebastian. I appreciate your calm style of making how-to videos, speaking clearly and explaining things to beginners, unlike other UA-cam teachers. I think other non-native English speakers will also appreciate this. It’s great that you provide good settings and show things thoroughly. I was looking for an answer on how to use the .csv style you made in Photoshop-plug-in and I got the answer.
Which do you prefer to use: Photoshop use or a local web browser? Please, make a video where you demo the good/bad sides of both and give good workflow instructions.
Thank god it's finally in photoshop. It's perfect for me so I can get down to editing and using in my Visual Effects pipelines. I absolutely needed this. Thanks!
post some of the outcome, please. I love PS actions. Maybe a combination of both things will be awesome.
This plugin is 6 months old at least...
Thank you Sebastian, another great tutorial, love it.
Happy to help and thank you!
the woman maybe part of doil i'm guessing... fantastic, thankyou!!! i couldnt figure out outpainting for the life of me :D
Will the prompt accept Textual Inversion keywords and Lora inside PS as well? This looks very useful, and basically what I think Adobe should have done to begin with instead of adobe Firefly
This plug-in does accept Lora
It's amazing. Does the installation from this tutorial work on the Mac?
I hope that you will make a video on the installation method step by step, I will be very grateful to you
will it run on affinity photo ??
Looks like this will be part of my toolset especially with a few of the glitches sorted. Props to Abdullah Alfaraj for developing it.
What do you think about Photoshop using generative fill instead? I find that it copies the existing style without asking. Which I think is amazing.
It's very good but almost always leaves a line where it fills
@@sebastiankamph If you are talking about the line around the edge or the file, all the pixels are filled up when I zoom in. So it will not affect printing. Or if you mean the seam between the edge of the square.
For example, I change the canvas from 1024 X1024 to 1024 x 1920. When I select the empty areas, I also select a little off the edge of the square. I don't see the seem at all. Only the edges of the entire canvas, which is an optical illusion.
I asked because I find that Ai generate actually picks up the style and make something similar much faster than SD. night and day difference.
It took me 15min to add details to an artwork while it took me an hour to inpaint and generate each graphic. It took me hours and still couldn't achieve what I did in 15min using PS.
Over 20 yrs on PS and 3 weeks on SD. I may be miss something though.
I have been using generative fill in photoshop for a little while and I've noticed it does a really good job of matching the style but, it's often in lower quality than the original. I can't get the generation as crisp as the original image. Maybe I'm missing something?
@@TheOriginalOrkdoop and then stable diffusion can upscale it again.
@@ekot0419 ooh! I'll try that. Thank you
Many many thanks for the clear Installation instruction haha...
Except when I setup my canvas and select 512 X 512 and hit generate. A dialog box pop up and said 'The command "show" is not currently available.
Hahaha... So close yet so far. Now I don't even know how to fix? Have you got any brilliant idea?
I had the same issue. Does anyone know how to fix it?
for me the outpainting do nothing ....also when i do it with the mask like you ? is there a bug ?
what is that nextml plugin beside it?
I got it to work with stability matrix + forge but inpainting just isn't as good as it is within A1111 itself. Also my Auto-Photoshop-SD looks different than in this video. Is there an updated tutorial or is this plug in just not very good? I am also wondering if anyone has found the solution to inpainting that lets you erase within A1111? Is there an extension or something for this?
That constantly appearing woman had me in stitches... then the healing brush solution i just lost the plot 😂😂😂
😅
Thank you for helpful videos. There is so much to learn these days. I would like to know how I can use your styles from the.csv file in this plugin?
cool stuff!
It really is! I love how my many years in Photoshop can sync with my newfound love for generative AI.
@@sebastiankamph i completely agree, integrating photoshop with the usual stable diffusion workflow makes some things much easier to deal with.
Thx Sebastian. Will it work of i run automatic 1111 from colab ? regards
Many female heads! I love it 😜🫶
Haha, thank you! Going early for biggest donor April 😘
I want to ask you, my friend, is this add-on free? And do you work on Photoshop 2019? Thanks
Free addon. I use latest version, 23.
Can this work with the Vlad version of SD?
What specs are needed to run this?
what is the "prompt shortcut" checkbox for? there is a severe lack of documentation..thanks for your great tutorials! this is a very good alternative to Invoke
I guess that would tell the plugin to use 'prompt shortcuts' or not (I've always left it on so I guess that's it?)
For those wondering, prompt shortcut in this plugin is the same thing as the 'Style' dropdown menu in the Automatic1111 UI where you can save and quickly apply prompt presets.
This is found in the Settings tab IIRC.
My wish came true, on April fools day.
Is this for Photoshop version 2023 only (or maybe version 2022 is ok, we just need the correct plugin)?
And when you said stable diffusion installation, do you mean Automatic1111?
Thanks.
didnt worked for my Photoshop CC15 Version, but now i get the 2023 and it works
and yes, by stable diffusion he means Automatic1111 :)
What about the VRAM ? how much VRAM will be used by photoshop, or SD will use the VRAM allocated for photoshop ?
The VRAM is not the main concern. Photoshop can simply crash when it will eat up all remaining RAM. Automatic1111 will keep running tho. I manage to use this on 4GB VRAM with more or less pain.
The Woman is the ghost in the machine LOL
Would be great if it is implemented in GIMP and Krita !
I saw krita plugin on list in ui
it is already implemented in Krita-->ua-cam.com/video/h-BIC9iuPGc/v-deo.html
How to attain a Horde Key - I see you have a password key in place - where can one obtain a key ? Ty
how can i use my API? it says i need to create a local network or something.. anyone have a tutorial for this maybe?
9:32 it's a scene from blade runner, there is always a large woman hologram in all these street scenes.
Makes sense. Thanks
what's your GPU ?
Rtx 3080
@@sebastiankamph ok, more than 8GB of VRAM certainly helps running Photoshop and SD simultaneously ..
@@anianait yes, at least 12gb is recommended for a smooth experience
@@jeethechamp7458 I have a 307Ti (8Go) in my computer, I have another 3060 (12Go) sitting around ... I am tempted to switch back to the 3060 then ...
*Adobe Firefly will add this soon*
Firefly is still in beta, and I have a feeling they're going to charge extra for it once it's released.
@@stagnation293 Adobe said it was going to integrated with all their CC Cloud apps such as Photoshop.
@@ZKatt as you might know now, you only get a certain amount of generative tokens. Each time you press generate you use one token. If you have all apps on your subscription you get 1000 tokens and from there it’s 500, 250, 100 etc. depending. If you run out you can buy more tokens.
Firefly has a crap ton of limitations. This is better.
Hi! How do I apply seg images to controlNet? or media pipe face?
Krita and I believe Gimp also have something like this, but when I last looked at the repo for them, the projects looked dead and hardly get updated which is sad.
I would love for more tools to implement them. But it's a lot of work. Let's see how it develops.
Thanks!
Current v1.2.3 version does not seem to be working. It connects to SD and gets LORAs list etc, but image generation just posts errors - Adobe Photoshop "The command Show is not currently available" and Adobe Photoshop "The command Set is not currently available". Is there any way to get the earlier 1.2.2 version? Other comments have been that this one still works.
Photoshop + Stable diffusion? ❤❤
Hello,
I have this error when I try to use Inpaint or Outpaint...
TypeError: '
New new new, i just got in and its pretty overwhelming lol. I just scratched the surface
Can you have Photoshop installed on one computer and Stable Diffusion running on another with this method? (and somehow connect them)
In theory yes, if you can get inside the plugin and change the ip for the api connection.
@@sebastiankamph Thank you, Sebastian! I finally was able to install Stable Diffusion so will go through your other tutorials. And hopefully can connect it through API.
This is amazing!
I have a problem with using contrlnet in photoshop , which is nothing after it is generated😅
Could this happen if I had installed SD in m1 MacBook?
WORTH NOTING! -> "The minimum Photoshop version that the plugin supports is Photoshop v24"
This plugin a little bit confusing to use comparing the old version.
Why you don't use Inpainting Models? If you use Inpaint Models, you will not get the edge problems when Inpainting.
When I try to generate an image, I'm getting the warning popup "the SHOW command is not available" ??? Anyone have the same issue or know what the issue is and how to address it?
Is this possible to do with something as old as cs2? I'm not paying an overpriced sub service...
15:54 Using Euler a sometimes make this color problem. i forgot how to fix it the other day. xd
To many errors for me to get this running. xformers and my cuda version is a bad mix.
jag har antagit att du är svensk i typ flera veckor nu, men måste fråga, du är det? :D
This plugin says SD not launched with API even when it does. Another plugin works, - this cannot contact with SD.
Plugin can't communicate with automatic1111, you forgot to add --api flag to the webui commands flags? i added it but it does't work. Why make a video without DETAILED instructions UGHHH
You only need to add --api like I showed you. Feel free to ask for help in the Discord with your screenshots.
thanks for sharing but please show the instalation steps next time
Wonderful, how humans take over AIs and turn them into toys.!!! Homo sapiens did it again! 😂😂😂😂😂
We're good at that! 🌟
So cool
Right!? I'm loving the new tools 🌟
inpainting mask doesnt work well
Honestly, I could not install it, and you did not perform the installation process practically
👋
Bob Ross like commentary 😅
Oh yeah, a licensed doctored.
glitchy..
woman appear sometimes, it's depends on the model, I think.
Maybe it tries to re-create Bladerunner 2049 hologram girl?
Am I the only one with a completely dif interface???
Why dont you use "woman" in the negatives if you dont want a woman?
Hey Sebastian, Sir, please make more semi-nsfw, semi erotic SDXL, CONTROLNET content please 😢
Might have better luck with content aware fill
Doctered image😂😂
no no no...
what a sad world we live in. i used to like digital art but i guess thats an end here. i guess the only art that still exists is the one drawn on paper with actual color you can see dry.
imagine a world were ever time you see a picture or text or youtube video or whatever you need to ask if its made by a human first...sry but i m not a robot and humans letting robots create something is no art
Inevitable.
Adapt or get left behind basically.
been doing that with a better plugin for quite some months now.
I'd love to know about a better as feature rich. Been testing many.
Called....???
@@thebundok my answer just crashed.... for some reason
i.imgur.com/Ah4AhEm.png
here is my answer in screenshot, because youtube makes all my answer crash for some reason, maybe a word is censored here, I've no ideas what's happening...
@SHADOW.SUPPLY the plugin is STABLE point Art ( not with the point )
Works with the cracked photoshop 😆
Does it require 400 gigs worth of custom models on your hard drive... Also yes!
😅😂
Honestly this plug-in is very very buggy, I've never ever gotten out painting to work and in painting is hit or miss at best. I've seen the author of the plug-in do some pretty amazing things with it but he's not offering any explanations for how he does these things and there appears to be almost zero documentation
I also found it extremely challenging to find any info on this. The outpainting works ok except I haven't found a way around that weird masking issue. The inpainting is pretty good as long as you keep the selection square.
@@sebastiankamph do you have any place that a person could send their art to you? Like a discord or a some kind of group
@@aegisgfx discord.gg/sebastiankamph
This is way too buggy Stable arts A I is much better than this one