Adding to Cart…
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.You currently have no notifications.
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
I like to refer to it as cookie cutter art. ;)
A fun tool. That is pretty much it. Might use it to create some background images.
It does seem like integrating 3D modeling with AI image generation is a possible solution to AI's problem with character continuity across multinple images. Solving this problem is probably the major driver behind this project, as hinted at in the PR statement. I wish the developers luck pulling it off.
I think so too. I like DS AI.
Hai...say wot iz the pointz of this tool?
I iz attach 2 samples using Victoria 4 in a simple pose, rendered in studio and run through the AI platform I currently use
Iz rush this off soon as I gotz home...so no background etc
One iz Photo realistic nd one uses an Anime setting...sure the handz iz need a little work but...Victoria...4...lookz pretty neatz huh?
Wot I iz expect from a Daz AI iz an interface that helpz me imporve my pipeline...an AI director of cinematography wot iz interface with my entire daz library of stuffz...not just G9...and allow me to render and tweak scenes using my stuffz nd promptz
Iz not get why we iz getting another 2d AI generative enginez when we iz a 3d community???
Good - It works, Bad - It took a long time. I waited a long time and decided to go off and do something else. I will try again.
I am using a time during waiting to think about what I would create next.
When is the touted image2image feature going to be released?
Quick feedback about options I missed most:
- automatically adding the seed number to the image name (or creating custom file names in general)
- a checkbox to automatically export the prompt as .txt file with the same name as the image
- advanced: include the whole workflow in the saved png. Drag the png back into the web UI to restore the workflow including prompts, seed, node settings etc.
Maybe I missed the instructions somewhere, but how do I access the advanced features (like controling size and aspect ratio) ... I paid for the pro version, but I only have the same controls that free version had.?
Does it say you are on thje paid fopr option at the top-left of the screen? The available options should be under the figure selection on the left.
Yes ... the top left says I am on the premium plan ... but I still only have the controls the free version showed. I will try logging out and back in while I await answers.
EDIT: maybe I am expecting too much ... many features are not yet available. However, are backdrops and lighting supposed to be available options yet ... the page description does not indicate if they are there yet?
It should look like this
Okay ... that is what I have so far. I was expecting too much apparently.
Yes, some features, such as inpainting, are not yet available.
I'm not sure what the point of this tool is.
You can, quite literally, do this by running a bunch of DAZ renders of characters you have through kohya & Automatic1111 or Comfy UI on a remotely modern graphics card, except you have a lot more control, it works much better, much faster, and it's free. And those features like inpainting, upscaling, etc... are already in there and implemented far better.
Also, DAZ is a 3D app... Even if it were any match for the free & open-source tools already available, this only applies to renders, doesn't it? Any interactive/game content, this does nothing?
I mean I get it, get on the AI train if ya want, but there's AI 3D generation, which would at least make more sense to incorporate into DAZ.
I think a lot more could be done if you guys released the checkpoints / lora's that were made for this, I know I could certainly get a bit more done with it
LoRA strength
Are there plans to add a dial to adjust the strength of the Victoria 9 and Michael 9 LoRA?
Customers might be able to achieve interesting results if they could manually adjust and experiment with LoRA strength.
Potentially unwanted side effects of LoRA
LoRA strength sliders could improve the overall image quality when using LoRA.
LoRA tend to orverride the style of the base checkpoint with sometimes unwanted side effects.
Example:
The Victoria 9 LoRA may not only add the face of Victoria 9 but also an auburn hair color, the ponytail hairstyle, clothing, monochrome backgrounds, lighting, etc. of the images used in LoRA training.
Lowering LoRa strength to 0.6 - 0.8 could provide the benefit that only the most trained attribute, in this case the face of Victoria 9 is applied.
###
Maybe LoRA strength sliders may be implemented when additional LoRA for clothing, hairstyles, lighting, environments are added?
I agree that right now it's kind of lacking in some respects, but it's a Beta. They probably want our input.
The world is changing. Especially in the art and music realms. I think DAZ is smart to hedge their bets and get this AI stuff going, because it's not going to stop.
If they can build it to work with DAZ Studio itself as an assistant to scene generation, or a post-work processing tool, then I think it might be nice to have.
I've been using Dream Studio by stability.ai and I really enjoy it. But you have much more control of the flow with DAZ Studio. And AI image generators do not create 3D characters for export into video games as far as I can tell. If I'm wrong on that one please let me know.
The game studios are going through the same chaos right now with AI. I don't envy them.
When the dust settles, who knows where we will be with all these new AI technologies.
We need the abilty to make higher resolution images.
Hello all...
I wish I could have been on the beta team of this but I wasnt...
I agree on the ability to make higer rez pictures with options..
With generative AI, higher res images are made via "upscaling". This can be more than simply increasing the image resoution. It can actually add new details, etc.
- Greg
ETA: Should have added that this option is typically included in most WebUIs for generative AIs.
This is a beta product, so: Welcome to the beta team!
I dont know what the fuss is about? I use the free Ai tools and Blender for Artists and the results are kinda blah. I get more pleasure sitting on my art desk with some plain paper with pencil and something I 3d rendered quick as a guide than Ai doing if for me. Just a suggestion. People take some art classes and just use quick renders as references for your art work.
Would it be possible to add the option to enter the Prompt Strictness as decimal number?
In theory you could recreate the exact same image in Daz AI Studio when knowing the prompt, seed and other used settings.
The Prompt Strictness does not seem to affect the result that much.
Still, knowing its exact number instead of a guessed slider position could help in some cases.
It seems that Daz has missed some important features and chances to improve beyond the other AIs. Ideally intergrating the AI into DAZ Studio so you could generate AI backgrounds or hair or costumes would have made it stand out.
I am experimenting with using AI as part of my workflow but I take one or more of my own renders into DreamUp as a image to use as part of the prompt. The option to use your own renders to prompt the AI means that yor composition and general lighting are kept and you can use the underlying render to correct issues (hands, extra faces etc).
This is a beta, and it isn't even a beta of the final product.
Richard, I appreciate it's a beta but DAZ still needs to listen to constructive criticism of it's features or lack of them. There are many viable alternatives to Daz Ai Studio to choose from and it's features will determine whether people choose to use it.
My point was that it isn't featue-complete, certainly not in regard to the long term text to 3D tool that was announced last year.
The main point for me is that it is coming from Daz 3D company
and I like it already very much.