[Stable Diffusion] Prompt Sharing and Learning Thread

[me3]

I wonder if the AI is hinting that we should be on the look out for something...
(hint: filename)

 

[Mr-Fox]

do you have an example image of the type of thing you're after? obviously doesn't need to be anything generated, just to have something to ballpark it

I meant in general. When ever I have attempted any widescreen images, it often or almost always results in either multiple subjects or the subject is lying down. I'm working on something special that would be awesome in widescreen.
Posting very soon. Maybe minutes..

 

[Mr-Fox]

This is using the awesome prompt of the eminent Sharlotte. Since I first saw these gorgeous images he shared in this post, I have thought of trying it out myself but with my own Kendra Lora instead of his Patricia TI. I have left the prompt mostly intact, only made a few additions and changed the settings to my own liking.
Thank you very much Sharlotte for sharing this prompt.

The original image:


My rendition:

 

[devilkkw]

I would be curious to try it for the benefit of making widescreen images. Can the extension in A1111 do widescreen? And with hiresfix? With the subject standing? This would be the ideal.

I'm understanding now the extension, but seem you have better control on sampler, preprocess and postprocess.
for hi.res i don't fund the hi.res option, but there are mane Upscale option, i think is these option (also in upscale you can use different model). I like auto memory managment, going up to in resolution and it automatic use tiled vad if you run out of memory.
For high size i got same problem as a1111, multiple subject.

 

[sharlotte]

Mr-Fox you are very welcome. Good to share the original prompts so we can learn off each other
On a different note, I see Sebastian Kamph has just released a 12' video on Comfy UI - very informative, especially if you have never used this type of flow:

You must be registered to see the links

 

[Mr-Fox]

I'm understanding now the extension, but seem you have better control on sampler, preprocess and postprocess.
for hi.res i don't fund the hi.res option, but there are mane Upscale option, i think is these option (also in upscale you can use different model). I like auto memory managment, going up to in resolution and it automatic use tiled vad if you run out of memory.
For high size i got same problem as a1111, multiple subject.

I ran out of reactions. So I say thank you for sharing your testing. Heartemoji

 

[Mr-Fox]

Mr-Fox you are very welcome. Good to share the original prompts so we can learn off each other
On a different note, I see Sebastian Kamph has just released a 12' video on Comfy UI - very informative, especially if you have never used this type of flow:

You must be registered to see the links

 

[Sepheyer]

I meant in general. When ever I have attempted any widescreen images, it often or almost always results in either multiple subjects or the subject is lying down. I'm working on something special that would be awesome in widescreen.
Posting very soon. Maybe minutes..

Here's a workflow file for ComfyUI to do a 2688 x 1536 widescreen using latent composing.

Takes between 20 and 30 minutes on my 6gb card depending on what other windows / apps I have open. Now, I haven't experimented one bit how to make these look better.

But! I think this is actually a deadend, and one should go with OpenPose control nets instead for widescreens. Imma try to do one soon(c).

 

[Mr-Fox]

Here's a workflow file for ComfyUI to do a 2688 x 1536 widescreen using latent composing.

Takes between 20 and 30 minutes on my 6gb card depending on what other windows / apps I have open. Now, I haven't experimented one bit how to make these look better.

But! I think this is actually a deadend, and one should go with OpenPose control nets instead for widescreens. Imma try to do one soon(c).

View attachment 2771940

Thank you! Heartemoji

 

[Synalon]

Is it possible to use tiling in controlnet to upscale the image without affecting the image quality?

And if it is can somebody tell me how please.

 

[Dukez]

Comfy UI inside a1111 is released. check

You must be registered to see the links

. View attachment 2770004

That UI looks way worse than the default experience you get lol, I'm assuming you can add on to it but if not that's an entirely pointless design IMO.

 

[me3]

That UI looks way worse than the default experience you get lol, I'm assuming you can add on to it but if not that's an entirely pointless design IMO.

Pointless, no. Different, yes, and much more "open" in what you can do, but YOU have to do it.
One is a UI made for simple usage basic and some more advanced things that's aimed at everyone, but you're limited to those options.
The other it's much more up to you. Like in a1111 you can't really "chain" rendering one image on different models, as an example.

It'll suite most needs just fine and do the job in most cases and cover ppls need and "introduction" to ai generation.
I guess an analogy might be, a WYSIWYG editors/software work great, but sometimes you need or want to write the "code" directly

 

[me3]

I meant in general. When ever I have attempted any widescreen images, it often or almost always results in either multiple subjects or the subject is lying down. I'm working on something special that would be awesome in widescreen.
Posting very soon. Maybe minutes..

Widescreen with little/no repeats
Single subject
Standing subject
Job done

 

[onyx]

Is there an easy prompt or a LORA for mirror/shower/steam writing?

 

[devilkkw]

That UI looks way worse than the default experience you get lol, I'm assuming you can add on to it but if not that's an entirely pointless design IMO.

the image is taken on default open, you can add-remove option by right click, and see there are many option to check.
I don't have official comfyUI and this is a simple way for me to check how it work. Maybe in future i switch to it, but actually i like to much a1111, or to be honest i have better experience on it.

 

[Mr-Fox]

View attachment 2772649
Widescreen with little/no repeats
Single subject
Standing subject
Job done

I forgot to say that I have managed to get a widescreen image with the subject standing and without any duplicates but the subject were always in the distance. I achieved this with an extension with regional prompting that I tried. Imagine a glamour shot photo of a beautiful woman walking down a tropical beach in a skimpy bikini, now this would be awesome in widescreen but you wouldn't want her to be far away. I have not found a way yet to achieve this without doubles or the subject lying down or being far away or any other weirdness. It's not a specific image that I'm after though but a composition.

Example:

 

[Dagg0th]

I forgot to say that I have managed to get a widescreen image with the subject standing and without any duplicates but the subject were always in the distance. I achieved this with an extension with regional prompting that I tried. Imagine a glamour shot photo of a beautiful woman walking down a tropical beach in a skimpy bikini, now this would be awesome in widescreen but you wouldn't want her to be far away. I have not found a way yet to achieve this without doubles or the subject lying down or being far away or any other weirdness. It's not a specific image that I'm after though but a composition.

Example:
View attachment 2773257
View attachment 2773267

have you tried using controlnet inpaint, to outpaint a image?

 

[Mr-Fox]

have you tried using controlnet inpaint?

View attachment 2773377

View attachment 2773387

Awesome!
No I have not messed with control net at all. It's on my to do list for sure. Thank you this looks very promising.
How is the workflow? Do you render an image with txt2img first and then use the control net inpaint?
As everyone has already heard to death, I'm very fond of hiresfix. Is there a way to incorporate it in the process somehow?

 

[Dagg0th]

Awesome!
No I have not messed with control net at all. It's on my to do list for sure. Thank you this looks very promising.
How is the workflow? Do you render an image with txt2img first and then use the control net inpaint?
As everyone has already heard to death, I'm very fond of hiresfix. Is there a way to incorporate it in the process somehow?

Yes, you render your image first from txt2img, then img2img.

I'm also very fond of using Hires.fix, but in this case it's not possible, because you have to render an image twice as big as the default 512 or 768, which most models use.

first create your image with a simple prompt:
"glamour shot photo of a beautiful woman walking down a tropical beach in a skimpy bikini"
512 x 768


After that you send it to img2img:
Put the image in controlnet an activate inpaint using impaint+lama as preprocessor, it's very important to put resize mode to "Resize and fill".

use a denoise between 0.75 to 1 and use your favorite widescreen size, in this case a 16:9


This will create an image from 512x768 to 1368x768 (16:9)

that is why is not possible to use Hires, because it will try to create from a 1024x1536 (if using x2 hires) to 2048x1536, it will take forever to render and the result will be terrible, I tried.

Probably you will have to try a few times before you get a satisfying image,

this was my first try using denoise 0.75.. result was not good

the final one I use a denoise of 0.9, but still have to use inpaint to fix any error

After that, you can use your favorite upscaler and do a final inpaint retouch

 

[me3]

Outpainting is most likely the way to go IF you have a background that allows for it. If there's specific patterns or elements you might need to do very small "additions" each time, with more general "easy" backgrounds you might get away with doing almost full image additions.

Previous image i posted was kinda meant as joke as i knew it wasn't what you were after but it was an easy generation that fit the conditions (hence the smiles).
So i kept these two on hand as the basics are in the prompt, you just need to work with it a bit more