Stable Diffusion plugin would be great...
Magic Man
Posts: 35
Given the recent Stable Diffusion plugin for rendering direct in Blender, think of the possibilities with an SD plugin inside Studio...
Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
I would think one would want to go the other way, apply DAZ faces and hands OMG the hands to Stable Diffusion generated images
there is already a program for that with a DAZ bridge BTW
it's called Adobe Photoshop
someone has also created a Stable Diffusion plugin for it
I too think that a Stable Diffusion plug-in for DAZ would be a fantastic idea. That way we can use DAZ's cameras and lighting to compose our Stable Diffusion images.
I love Photoshop but truly see no advantage to the PS Stable Diffuse plugin.
I find it much easer to render an image via Automatic 1111's local install of Stable Diffusion then export the image as a png or jpg, and open the image in PS.
I love the Blender plugin for Stable Diffusion as I can take advantage of Blender's camera and lighting.
BUT I have too many issues with materials when I export a DAZ model and materials and import them into Blender.
The Blender plugin for DAZ is a nightmare to install unlike other Blender plugins which install easily. The Blender plugin for DAZ is simply too buggy and cumbersome.
Until we can render hi rez animation frames in those illustrious SD styles in Blender,
the Plugin for it is utterly worthless.
I stand corrected. I decided to try the Blender plugin once again and it installed flawlessy in both Blender and Daz. The version I tried before I had serious issues getting it to install on Blender.
I do not do annimations. I do still images. And I think it would be a great asset.
Photoshop sounds like a good software for altering 2D images made with SD. You can add lights and things like that with it, and it's got an entire library of filters available for it.
As far as I know, all SD does is scrape other peoples images from a list of web galleries then layer them together, apply filters and a color palette to them then merge the layers.I'm probably wrong, but I don't see it doing much more than that as it doesn't actually create the original seed images.I get this idea from the posts and news stories I've read about it using watermarked and copyrighted images in it's "creations" that it's farmed from those galleries.
I'm not seeing the advantage of having a plugin for a 3D app like Studio or blender for SD or any of the other image scrapers. They don't use 3D assets to create the images. Just 2D images it's downloaded from public online galleries.
You do realize that the Photoshop Neural filters are using AI. Many of Topaz's new filters are using AI. Corel Painter and Corel Essentials are using AI for some of their auto painting sets.
Stable Diffusion can also be used to created textures for 3D models you are creating or texturing in Blender.
Sorry, but like it or hate it, AI is here to stay. The only AI I do not like is for autonomous cars..
I wanted to create an image of a scarecrow which has arms comprised of a hockey stick.
I did this sketch using a pencil and paper and scanned it into my computer.
In Photoshop I colored his various elements - hat, hockey stick, shirt, etc.
I used the shaded image I created in Stable Diffusion using im2img and after many interations, I got this..
How can anyone claim that this is not my creation?
When I google for "scarecrow comprised of hockey stick" I get only 3 or 4 images and none close to mine.
I wanted to create an image of a scarecrow which has arms comprised of a hockey stick.
I did this sketch using a pencil and paper and scanned it into my computer.
In Photoshop I colored his various elements and added some noise - hat, hockey stick, shirt, etc.
I used the shaded image I created in Stable Diffusion using im2img and after many interations, I got this..
I did flip his stick via photoshop so the stick blade is facing downward.
How can anyone claim that this is not my creation?
When I google for "scarecrow comprised of hockey stick" I get only 3 or 4 images and none close to mine.
I started to sketch some images of kids playing hockey and a few car / hot rod design with intent to bring them into Stable Diffusion.
I've not decided if I like it or not yet.
I would be impressed if I could import a human female mesh and tell the AI "Make a photo real texture set for this Female" and have it do exactly that.
Currently, to me anyway, it just seems like a glorified scraper script of which many have existed before without the AI in the name.
---
I see it like the current "Drones".
We've had these since I was a little kid, and I'm in my 50's now.
Back then we called them remote controlled airplanes, helicopters, cars, etc,etc.."RC" for short.
I remember my friends and I would put model rocket engines in the backs of the cars for drag racing them.
Used the head light switch on the controller to fire the rocket when you wanted a little extra (well alot) boost to put you out front.
There was no burn control, so you got the whole fuel load.Some guys were experimenting with cutting the engines down, removing fuel from the tube, things like that when I moved on to other things.
Many expensive for the era RC cars were melted down on the track after spectacular crashes (the street out in front of the house :) )
The hobby saw a slow down at or near the end of the 1980's.Once the name "drone" caught on, it saw a huge uptick though.
AI might be the "Do it for me" button for the art world in future, but I've not seen it do anything really impressive yet that I would call Artificial Intelligence.. Of course I don't follow the industry nor do I experiment with the software. I just see the results posted in the galleries here, at Rendo and also deviantart.
Maybe there is one of these AI's that can texture a mesh for me at my command and deliver what I asked for, I've just not seen it yet.
Sorry IceCrnMn if you are undecided whether or not you like my image.
You are a youngster compared to me. I'm in my late 60's and will be 70 in a year or two. LOL
I worked over 35 years with computer technology (database programmer, systems analyst, LAN administrator, etc) and never seen AI that is coming down the pipe like this.
Cheers
No need to feel sorry, and I do appreciate you engaging me in conversation about the subject.
As I said, I'm just skeptical of the AI part. Not the first time either of us have been told that some buzzwordy software was going to change the future.
As for liking the images SD currently produces, I do like some of them,but nearly all seem to have a specific style to them.
I assume the software is recording "successful" choices, the ones people are choosing to keep instead of rebuild, and applying the same logic patterns to subsequent requests.
How Vladimir Chopine uses DAZ3D and AI generated background.
If Vlaidimir says it's good enough, than it's good enough for me... :-)