Will products already owned work in AI Studio?
alaltacc
Posts: 151
I believe they will not, as probably this wasn't covered by the initial license. As a Stable Diffusion user, I'm not really optimistic with AI Studio because of this added cost. The problem is not the $3.99 a month subscription fee (this is affordable considering other online AI image generation services), but the need to buy everything again compared to the tools (LoRas for consistency, Controlnet for posing and scene composing, dozens of models, etc) that are available for free using a local Stable Diffusion installation.
That said, it's a good thing seeing Daz trying to include AI into the product, since this is clearly the future.
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
No. Looks like they want to make you pay for it all over again.
Frankly, it's ridiculous for them to think I'm going to purcase characters when Stable Diffusion is open source and I could easily make my own LORAs with the characters I already own.
I disagree completely that it's a "good thing." AI images already exist - Daz should remain seperate.
Theoretically you CAN'T create LoRas out of the characters (the license doesn't allow it). But you CAN create LoRas out of a lot of other things, including photorealistic images (that apparently Daz AI Studio is not capable of creating - the images look like Daz renders), so your point stands.
As the Daz being a separate thing, I believe this is what they're trying to do: there's the "classic" Daz 3D on your computer, and there's Daz AI Studio online. The question is: how long can Daz 3D survive with it's business model since Stable Diffusion renders are free, faster and have better quality in general? There's the question of continuity/consistency (it's not possible, for example, to create several renders of the same room in different angles using Stable Diffusion for now, but it's trivial to do that inside Daz Studio), but this is something only a fraction of the users really need.
Daz should play to its strengths. I use Daz Studio to create comics and that's impossible to do with Midjourney or Stable Diffusion. Trust me, I've tried!
The consistency isn't there for clothing, sets, lighting, hair, or anything else.
As more people get into art (maybe through Stable Diffusion) they will want to tell stories. Daz could have positioned themselves as the "next step up."
Their best bet would've been to somehow integreate SD into Daz Studio to improve a render after it's done.
I couldn't have said it better
I agree with you. I've posted about it in a (long) thread abouth AI in the forums: the one thing that is unquestionably better in Daz compared to any AI generator is consistency. There's no way to have perfect consistency in ANY generator, even using LoRas, Controlnet, specially-trained models, etc. AI art is simply based on randomness; Daz studio is completely deterministic and based in a virtual lighting model (for renders) and 3D models (for objects).
Put AI INSIDE Daz3D as a helping hand to create a scene, for example, and you will have something that simply doesn't exist yet. Like: "Victoria9 wearing DRESS1 and SHOES2, standing in room ABC, talking to MICHAEL9 wearing PANTS1 and SHIRT2, seen from above, low light" could pre-create a scene using these assets and pre-posing the characters, so you could do your adjustments from this starting point. Also, when rendering, Daz could give you an OPTION to use an AI rendering after-texturing model to texture your scene with the style A, B or C.
Is this simple to create? Hell no. Is it feasible? Hell yes. And it would REALLY set Daz apart from EVERYTHING else.
Or allow us to generate skies or backdrops based on the lighting we add to the character, or do something like Photoshop's neural filters and change a face expression or the eye direction or even retouch poke-through, etc. There are so many ways it can work with DS to make our workflow better, instead of offering another SD clone.
I agree, a full scene generation based upon AI is most likely beyond what is currently possible.
But they could aim to tools like background generation or modification, or improvement of lighting / skin tones.
For sure it is more work than training an SDXL model with renders (I guess this is what they did with AI Studio) but at least it would be something more original.
As far as I know they are still working on the Text to 3D tool they trailed last year, I'm not sure if this is a step on the way or a spin-off from it.
I'm curious what it'll become. Concerning buy again or not, i don't know, but i could imagine discounts or inclusion for stuff you own in 3D. The whole thing is beta and i bet (ha!), that there is a chance, they'll actually have to figure out what pricing model and how where when to use. It's probably or will be a (modestly) unique approach, so it is a bit of a jump i'd recon.
In the context of what is currently in this tool, I was at least hoping for the one thing other AI image generators can't do: stylistic consistency for characters. We also really need something like OpenPose or a pose library.
I'm mildly optimistic that they'll try such. Looks like you can already select characters, so maybe that coul become a thing. Probably the workflow for multiple things could become to render them individually, though one sometimes might need interaction.
Not sure this currently is within scope of any of those, but i could imagine a pre-text phase, which is a pre-generation text phase, where the text thing tells you what it relates to what, so you could get an idea about if a formula, a word or a (partial) sentence relates to something you want it to relate to. So that would be something like a grammar-setup phase, working only on text and somehow visualized context/s. Maybe that's another bucket of research...
WONDERFUL idea!!! WOW.. I'd use Daz more too... and I render something almost every day! lol
Yes it might actually have value if users had an efficient workflow within daz studio which meant they didnt have to leave to use AI. Having all the granular and fine-tuning options for Stable Diffusion available after render time would improve user experience for those that use AI.
Would it undermine Daz's sales though? Perhaps, and vendors may start selling simplistic clothing, characters, if users are expected to just enhance them in AI afterwards. Daz Studio then basically just becomes a posing tool.
I dont presume to know the way forward that would make everyone happy. Personally, it would make me happy if Daz just become a character creator for Blender, fully integrated into Blender.
From the FAQ: "Our Free and Paid users will have access to our core characters Victoria 9 and Michael 9 with other characters available for purchase from the store in the future that you can add to your account."
Sounds like another purchase, maybe like 3D print licenses or the meta-mixer tools.
This whole thig is a big, big, big no for me. It really makes me regret money I've spent with Daz, and makes me strongly suspicious of whether to trust or spend more money with this compay. As if the NFT mess wasn't enough of a warning...
You are right: consistency (stylistic and for the characters/clothing) is the ONLY thing that could give "Daz AI Studio" ANY advantage over all the (many) others AI generating services out there (and this is without even mentioning Stable Diffusion in a local installation). As much as they can advance this online tool, there's simply too much competition online and too many extensions for people (like me) that use Stable Diffusion in their own computers to deal with. Controlnet, LoRas, tens of different models, upscaling, and the list goes on. To REALLY get somewhere, they would need to beat Midjourney OR Stable Diffusion, and that's simply not happening IMHO.
Again, the way Daz could do something really new and that still doesn't exist on the market would be to implement AI INSIDE Daz Studio. Prompts for creating scenes (you describe the scene and give the names of the assets instead of picking one by one from your library, and Daz creates a scene for you to tinker with afterwards), post-rendering using AI to texture or style the scene, etc, all this OPTIONAL inside the existing Daz Studio (people can simply continue to work as they do today OR use the new features). Seriously, this is where I was thinking Daz were going at some point, not creating a new online image generator limited to a few models (probably LoRas they created out of renders of these models).
I've thousands of dollars in my current library. I'm disappointed they want to charge me again. :(
I would have preferred AI inside Daz. Having the abilty to get a jump on a scene with something that looked at what content you had (or even upsell to content you don't) would have been great. I'd rather have been able to say "Set up a scene with an astronaut running, and tripping over a rock, strongly backlit by the sun, but environmental lighting providing the majority of the light." THAT would have been helpful.
The image generation, while technically impressive, is super depressing knowing that I can take the time to create something and take hours, or someone can type a prompt and get 90% of the same thing with no effort.
I already have to provide proof where I sell book covers my art isn't AI generated. I'm glad for that stopgap, but man is it frustrating.
I honestly thought this was what it was going to be when I first heard the announcement. My disappointment fell off a cliff when I realized what it actually was. If we could select items from our existing library of assets and use AI within Daz to help stage the scene, plus some extra lighting finesse etc, this would be the ideal scenario. It also ensures people keep buying from the store and gives them an extra incentive to do so. It actually supports the current customer base and enhances the experience, instead of clumsily upsetting and turning away what appears to be quite a lot of people from the comments I'm seeing.
Don't forget that this is a beta - there is more to come (not that I have any details, but I do know that they are still working on/towards the text to 3D thing that was announced last year).
I like the idea of Text to 3D assistance in DAZ Studio (ver 5?). A good first step would be assistance with things like: adjusting a character's expression, posing the character in a scene, and tweaking the lighting and atmosphere.
Lol wasnt AI supposed to DeMoCratiZe art??? Looks like its just another new pipeline to siphon money off people with a passing interest in art in order to pay ever increasing fees to make it. Reminds me of streaming vs cable. In year or 2 the costs of these products/subscriptions will cost as much as having adobe or zbrush. And for what? an amalgamation of other peoples styles.
There is a free tier, though. Whether it will get more free content in futue I don't know, but it certainly wouldn't surprise me.
DAZ servers (even if they are rented servers) cost money (very costly graphic cards, electricity, etc). I can understand charging for the use. My problem is that it looks like we would need to buy every product again (or at least to buy an extension of the license) to use them in the AI generator.
I can understand that there are licensing issues even with the PAs (when they put their products for sale there was no provision for creating LoRas out of them), but for someone that already uses Stable Diffusion (a lot) in my own computer with all the LoRas, Controlnets and whatnots on one hand; and that have spent a good chunk of money building a Daz library to use inside Daz on the other hand (I'm even a Daz+ subscriber), it's really really hard to pay extra to use these (same already paid for) assets in the online AI generator. So, apart from my view that the AI shoud have been integrated in Daz Studio instead of being used as a stand-alone product based on Stable Diffusion (and vastly inferior to a local installation of SD), there's also the frustration of not being able to use the products that I already have without paying extra.
your music on cassettes -> cds -> mp3 s -> drugs:-))... pay up!
Has DAZ actually defined the usage price for their content to be used, will it be the current price, or will product currently owned be AI useable? Will they limit it to DAZ only content, will Artists' content be included? How about Artists' content we currently own for DAZ Studio? These questions have not been addressed to my knowledge and I don't believe they are sure either. It's pointless to get DAZ Studio users to rebuy products just to use the AI. That's just very poor business planning, and those that are assuming content has to be repurchased are already providing a clear backlash that will result in some very negative responses that would have DAZ board members fatting for someone’s head. If it were me, the best business model would be to make all DAZ original content already owned available, and DAZ artists' content, already purchased would have to be at the discretion of the artist as to whether they opt in for AI use. All content for use with AI in the store would be marked "AI Friendly". This would satisfy current product owners, allow Artist a new customer base increasing their sales, and allow DAZ to generate new revenue from AI users and increase the revenue from current DAZ Studio users.
Also, Image to image, and Inpainting are going to be more relevant for DAZ Studio users, in my opinion. We have the tools to create what it is we want within DAZ Studio, using AI may assist in adding details and improving on elements we may be lacking. Also, AI art improves with use, just as we improve with practice. I saw some pretty cold receptions to DAZ AI on YouTube, and it was clear that they really didn't understand how AI worked, and they were using the free version and did not Prompt well, so they got some real poor images.
Let’s get some confirmation on the business model first and avoid speculation. Depending on how DAZ approaches this moving forward, they may either make bank, or get tanked.
The only details released are in the Blog post etc.
It may be the future in the commercial world, where it's speed that counts, but speaking from personal experience, iI didn't find it a lot of fun. Ultimately, I didn't feel I was the one making the pictures. I didn't have direct control, I could only influence the outcome. It was a little like commissioning someone else to make me 1,000 images, selecting the best ones, then fixing that person's mistakes.
I played with DeviantArt's DreamUp briefly, until it told me that it didn't like a word in the prompt, and further breaches would make me lose my account. Running locally removes such restrictions, and is free, as you point out. And let's face it, we all have the hardware.
I'd like to have a play with this, though I certainly wouldn't spend money. It's still not clear to me what the TOS specifically for AI images are, but after my DA experience, I'm not going there.