If you get pleasure from it, and some people like myself appreciate the results and the work that's clearly gone into it, then I would say continue. Perhaps it is ithe high quality of the result that makes some people think that there must be a trick to it,
Definitly!! I have spent weeks putting together my original work with little pieces of AI work.
I use Chat GPT for advice. "What should I call this?" or "What would look good with this?" "What style is this?"
Oh, I just purchased an Apple Ipad based on her recommendations. I didn't want to work on an Ipad. I just wanted to watch movies and surf Facebook...things like that.
I didn't think I could get an Ipad for just that that didn't cost $600 or $700. I think it was around $350 or so...under $400 for sure :)
I agree with everything you said. I don't get defective AI renders, because I use a quality AI site (OpenArt.ai). I have to pay for every render I create; therefore I can’t afford to create 100 images of the same thing. I use a lot of DAZ 3D in my AI renders.
These are awesome :)
Yes, I used so much Meshy AI to create props that I had to pay again last month.
What got me started so much on using AI among other things, it that I have asked and asked where to get something used in promos or to make things (Which I have already mentioned.)
I have had Chat GPT to try to find an outfit. I have trained her to look at Daz3d, Renderosity, Poser Content...etc. I get embarrased asking you Daz3d folks about an outfit or hair, etc.
I asked her to find a really cute hair product:
Thought for 33s
I found the hair. It appears to be dForce Bixie Cut Hair for Genesis 9.
Best match:dForce Bixie Cut Hair for Genesis 9 at Daz 3D. It is the short bob/pixie “bixie” style shown with Lyna 9, and the product description mentions the bob/pixie blend, soft face-framing layers, optional sleeker pixie/undercut morphs, and two-tone ash blonde/silver/dark-root materials.
It is also included in the Lyna 9 HD Character Bundle at Daz, along with Lyna 9, Dark Web Dive Outfit, M3D Xyber Vision, and the hair.
PoserContent also lists it as dForce Bixie Cut Hair for Genesis 9, but its “where to buy” entry points back to Daz 3D.
I did not find the exact same hair for sale at Renderosity. The closest Renderosity direction would be searching for short Genesis 9 pixie/bob hair, but the exact match is the Daz product above.
If Sturgeon were alive today, he's have said 99.9999% of everything is crud.
..I had the wonderful opportunity to have met and speak (actually listen more than talk) with Mr Sturgeon at a small Sci-Fi con in Bellingham WA back in 1983 a couple years before he passed on,
That said
I have a background in traditional art that stretches a lifetime so I tend to think visually rather than with words. Hence I am terrible at composing prompts but have little issue "blocking" (a term used in theatre) a scene. Working on a scene in Daz is actually not much different than designing/building a set for stage or cinema and determining where the actors are to go (which is something I've also done.)
Granted with painting and drawing I didn't have to worry about, things like light temperature, lumens, SSS, refraction index and such.
If Sturgeon were alive today, he's have said 99.9999% of everything is crud.
..I had the wonderful opportunity to have met and speak (actually listen more than talk) with Mr Sturgeon at a small Sci-Fi con in Bellingham WA back in 1983 a couple years before he passed on,
That said
I have a background in traditional art that stretches a lifetime so I tend to think visually rather than with words. Hence I am terrible at composing prompts but have little issue "blocking" (a term used in theatre) a scene. Working on a scene in Daz is actually not much different than designing/building a set for stage or cinema and determining where the actors are to go (which is something I've also done.)
Granted with painting and drawing I didn't have to worry about, things like light temperature, lumens, SSS, refraction index and such.
What a thrill to listen to Mr. Sturgeon. A time that will always be with you.
You are one of the artists that facinates me. The animation is brilliant!! Please share how you did that!! It looks like Disney or Pixar...
You don't have much in the Daz gallery. Do you have more somewhere else?
Do you share your story? You have that God given talent that I have wanted forever!!
The animation isn't mine, it is a GIF I found many years ago (really love that film and glad they didn't ruin it with a bunch of sequels) don't do much with animation as I have an ancient system which just doesn't have the "horsepower". to handle something like that. I even hold my breath while running a dForce sim, as that sometimes crashes the display driver.
I do have a more comprehensive gallery over on DA I use the same name there. It also includes some very old drawings and sketches I did, the latter when arthritis had set in which eventually made it too difficult to continue.
I use AI with renders because I hate waiting. I render my scene in the pose I want and ComfyUI and Qwen Image Edit {Locally} lets me relight and change my scene however I want. I don't trust AI to place people in the position I want them in and the truth is OpenGL or Filament is enough to get me the renders I want. I create the content, AI relights and changes the lighting and skin and clothes and I can add thing and remove things with no problem.
if you think about it, is AI much different than buying a model from Stonemason, for example, or any other PA or DAZ, and use it in your image? You didn't create it, Its someone else's creation/imagination. There it is. Viola, art, I guess. Yeah. we can manipulate the image within bounds of the model being used. We can manipulate AI until we achieve what we want. Its all on the same path really. An advancement of human thinking nd creativity. Most of what we produce here with DS is based on formative art that has developed over centuries.
But, I don;t like AI and data gathering for the stress it puts on the planet and our resources. Even our bodies.
This is not a correct example.
Street Photographers don't create streets and still it can be art.
The difference is, that you can reinvent a street view and a Stonemason scene view and give it a new meaning.
Generative AI on the other hand presents a scene to you including camera angle, tone mapping ect..., the user's job is to just make selections.
With generative AI, I'd be missing the joy of wandering my camera through a 3d space in order to find that one perfect view to render,
finding that one perfect light set, that somehow resonaces with me and having that one pose and expression on my character, that tells a story.
No generative AI can do this for me. As an artist, I want to project my vision, not just make a selection on a machienes offer of results.
So, declaring generative AI scenes equal to already existing assets or real life scene, shows a lack of understanding the artistic process.
The difference is also, that humans process into art, based on human art work. That matters.
I don't have to respect, what a web trained machien makes of already existing human art.
At the end of the process, there is a human bragging with a result, created by a machiene.
In many ways, AI is like a skilled friend, that can be asked to create some artwork.
Y: "Hey Vincent, make me a nice sun flower field."
V: "Here you are, I made five different versions for you, which one, would you like to keep?"
Y: "I like nr3, but can you add a tree, please?"
V: "sure"
Y: "perfect"
Y: "Hey look, I painted a georgous sunflower field, it's art."
This is generative AI, nothing more.
Again condesing the thought:
I want to know, what a human does with the available resources, not AI.
BTW, I find Stonemason sets frustrating, because they are already so perfect.
The question is always, what to do with those?
There is just nothing left to do potentionally better.
I have a lot of Stonemason products and I almost never use them.
Here is Stonemason's starship bridge rendered in DAZ 3D, and a second image where I had AI convert that image into a comic book look so I could use in a comic I'm creating
Generative AI on the other hand presents a scene to you including camera angle, tone mapping ect..., the user's job is to just make selections.
This is generative AI, nothing more.
The two 3D renders that I am currently working on and editing with AI, each with 30+ layers and layer masks to allow me to paint in what parts of the AI generations I want to keep and manually postwork so that they look as I want them to would disagree.
Those against AI almost always think that all generative AI is "type some words and get a pretty picture". A lot of it is, and that's what we call "slop". Some of it most definitely is not.
I have spent a ton of money on AI digital art to enhance my art. I purchased these as digital scrapbooking kits. This was before I even understood what AI art was.
This is about why and how people use AI. If you don't then this isn't the thread for you, please don't try to turn it into a discussion of AI's putative evils or lack of validity.
@jan_scrapper: could you describe how you use AI to improve the draping of clothing? Are there limits as to what the AI can do in terms of clothing movement?
This is about why and how people use AI. If you don't then this isn't the thread for you, please don't try to turn it into a discussion of AI's putative evils or lack of validity.
Please re read the first post, you missed an important part:
Some of the reasons I use AI to enhance my Daz Studio renders:
1. I cannot get anyone to make the clothes or the props I want. (This is lately, I want to say. Many, many people made me custom clothes years ago. Mada, EvilInnocense, Roylo, and others. I thank you so much!)
2. I cannot afford to get all of the hair or clothes or props that I need for my project.
3. I don't have a ton of Photoshop talent.
4. My computer is 9 years old and NOT a gaming computer. Daz Studio crashes because of lack of power.
5. My eyesight isn't what it used to be. I'm 70 years old and sometimes working in Daz Studio trying to add several decorations to something is very difficult to get the angles, etc. right.
6. Many others, I just cannot remember more.
I use Ron's brushes to enhance my work, I use HDRI images for backgrounds...I use Chat GPT to make images of things to add to my renders.
I didn't put my work in order, however the renders without bling are mine done in Daz Studio.
I made the bows with Meshy AI after designing them in ChatGPT.
It is a blast to start getting props I need for a particular project.
Items used for Pretty Little Things:
MB Gier for G9F - Daz3d
Addy Florence for G9F - Daz3d
Charlotte for G9F - Daz3d
CHT Memol for G9F - Daz3d
VYK LA Pouf - Renderosity
Elegant Display Props and Scenes - Daz3d
DM's Composed - Renderosity (For lights)
The Swan - FK Design - Renderosity (For the ballerina skirts)
Chat GPT: Used for the decorations of the Pouf Hair Bouquets and the Pouf Bodysuits AND the Title Banner
Meshy AI: To create misc props to use as decorations
I am showing my drafts of renders...a few of many renders.
My PSP workflow had so many layers that I started getting really confused.
You might see many problems...double feet, double hands, etc.
Thanks for posting the before and after. Interesting to see how the AI also improved the bends of the dancers' legs (ie the calf is not clipping into the thigh).
@jan_scrapper: could you describe how you use AI to improve the draping of clothing? Are there limits as to what the AI can do in terms of clothing movement?
I told Chat GPT:
"isolate woman and her clothes and hat on a separate layer, correct her pants legs and her jacket arms, high resolution, transparent image"
I was still learning Chat GPT. Chat CANNOT do pure transparent png images. So, if I want something that is lace, that is a ton of postwork to get the seethrough item.
This is not a correct example. Street Photographers don't create streets and still it can be art. The difference is, that you can reinvent a street view and a Stonemason scene view and give it a new meaning. Generative AI on the other hand presents a scene to you including camera angle, tone mapping ect..., the user's job is to just make selections. With generative AI, I'd be missing the joy of wandering my camera through a 3d space in order to find that one perfect view to render, finding that one perfect light set, that somehow resonaces with me and having that one pose and expression on my character, that tells a story. No generative AI can do this for me. As an artist, I want to project my vision, not just make a selection on a machienes offer of results. So, declaring generative AI scenes equal to already existing assets or real life scene, shows a lack of understanding the artistic process. The difference is also, that humans process into art, based on human art work. That matters.
I have a course coming out in the next few weeks called Art-Official Intelligence. I will walk you through the parallels between using you own artistic expertise and skills- and how they 100% translate from Daz Studio to Daz Studio + AI.
I explain with numerous examples of (classic and established) Photoshop techniques that are 100% the same when using AI.
If I know cameras and lighting in the real world, I will bring that knowledge to Daz Studio. I will use the camera and lighting settings to enhance my Daz renders.
Seems like a simple concept.
I would also do that with AI- use my own knowledge and system, and workflow to translate it for the new tools available. That idea is heavily featured in the course.
I don't need to imagine an "average Daz user" or imagine an "average AI user" to make my arguments make sense. We can all insert ourselves and start from where we are.
Do we all know that you can give AI your camera type/brand, film stock, all associated settings, time of day, position of sun, intensity, and even choose processing for the film? Do we all know you can submit your (personal) photos and artworks for reference? Advanced AI work includes all the same settings that Daz Studio has and more?
Do we know that I can give the AI -images of the village I visited on vacation and use that for the settings or make variations to create a place that looks real, but is not?
It’s Daz Studio + AI for the win. It doesn’t make sense to trade in all the Daz experience, the library and knowledge of environment, lighting, posing etc for the option of typing a few sentences and hoping for the best. It seems more logical to add AI to your toolbox and expand on the art you already make.
I also detail in the workshops how important inspiration is to the creative process. Why am I rendering stuff in the first place? Daz Studio is inspiring. The vendors make stuff that makes me want to use it for my arts. That’s why you buy stuff- from an artist vantage point. Obviously, some of us have commercial concerns and buy the stuff they need for a client/project.
Anyone (and maybe that’s the problem) can prompt their way to nice illustrations.
Most Daz renders are nice illustrations.
Turning the illustration into art is still a very (amazing) human thing. The rules of/to art haven’t changed.
There ain’t no easy way to get to art. That’s the whole point of art. That’s what keeps art, art.
AI renewed my interest in DAZ studio. I'm mainly interested in realistic renders. But I need them to be specific. I can use DAZ to create the specific scene and then use AI to make it look realistic. Neither tool does what I want by itself.
I'm curious to find someone to provide a "Comy UI" for the dummy DAZ Studio user. I have decent hardware, but I haven't had much luck getting it to do much to improve my renders. I'm not terribly interested in using the big LLM chat portals. I prefer staying local for everything, which is getting harder and harder to do.
I'm curious to find someone to provide a "Comy UI" for the dummy DAZ Studio user. I have decent hardware, but I haven't had much luck getting it to do much to improve my renders. I'm not terribly interested in using the big LLM chat portals. I prefer staying local for everything, which is getting harder and harder to do.
Look up pixaroma on YouTube. That's where I got my ComfyUI start. If you have 16GB VRAM you can run (for example) 14b Wan 2.2 models with minimal swapping.
I'm partial to InvokeAI - I find it to be a more familiar interface for Daz and Adobe users, very image viewport centric. It has layers, masks, local guidance, and you can draw straight into it if you like. It runs locally and has a pretty good model manager for trying out differrnt downloaded models, LoRA's etc. The commercial side of the company was aquired by Adobe about 8 months ago, but it's (free) community supported tool lives on and is still being updated. Their YouTube videos are pretty good and give you a good idea of how the tool can be used.
I have moral and environmental and political objections to GenAI.
So did I, when I first learned about the tech three years ago. I spent the rest of the year learning more about it so I could better defend my position, only to wind up adding it to my own toolkit after all. As an artist with a strong, decades-long interest in photorealism, my objections became overshadowed by my need to express myself creatively, so I compartmentalized them and set them aside.
Incidentally, not having to rely on 3D content vendors and their ideas of women's fashion has been a godsend for me.
I have moral and environmental and political objections to GenAI.
So did I, when I first learned about the tech three years ago. I spent the rest of the year learning more about it so I could better defend my position, only to wind up adding it to my own toolkit after all. As an artist with a strong, decades-long interest in photorealism, my objections became overshadowed by my need to express myself creatively, so I compartmentalized them and set them aside.
Incidentally, not having to rely on 3D content vendors and their ideas of women's fashion has been a godsend for me.
I want to respond to everyone, however this last sentence really got to me.
I really don't think PA's read the Product Suggestions. Or, they don't read mine, anyway.
I have wanted a Dforce skirt with a ruffle in a circle so I could let it flow over anything I wanted. I still want that.
However, I have used ChatGPT and Meshy AI to get what I want.
I am NOT happy about it, but that is better than NEVER realizing dreams I have had for so many years.
I am working on a Broken Doll Repair Shop with a Steampunk theme.
I had Chat GPT make a Steampunk chest with extra arms and legs in it. I have Meshy AI turn these images into props :)
Then, I asked Chat GPT to make a couple of limbs to lay in front of the box.
They thought it might be considered bad.
You just never know :)
Try Grok. I left ChatGPT three months ago when it started rejecting more and more of my prompts - including prompts it had previously enhanced for me. I turned to Grok against my better judgement - I'm not a fan of the CEO - but it's been a helluva lot more useful, and has yet to reject anything I've thrown at it.
Comments
Definitly!! I have spent weeks putting together my original work with little pieces of AI work.
I use Chat GPT for advice. "What should I call this?" or "What would look good with this?" "What style is this?"
Oh, I just purchased an Apple Ipad based on her recommendations. I didn't want to work on an Ipad. I just wanted to watch movies and surf Facebook...things like that.
I didn't think I could get an Ipad for just that that didn't cost $600 or $700. I think it was around $350 or so...under $400 for sure :)
****************************************************************************
I'm getting so off topic. This is about AI art. Sorry.
These are awesome :)
Yes, I used so much Meshy AI to create props that I had to pay again last month.
**********************************************************************************
What got me started so much on using AI among other things, it that I have asked and asked where to get something used in promos or to make things (Which I have already mentioned.)
Please make more cute retro clothes!! - Daz 3D Forums Has 1 view on no comments
Looking for gold, small leaf body art and the character for HS Mirage Strands Hair For Genesis 9, 8, - Daz 3D Forums After looking for what seemed forever, I came to the conclusion that the small flowers and leaves and vines, gold body art was either AI or something the promo artist did and is NOT for sale.
Can you help me find this blue and white dress from Aureliia Doll Hair? - Daz 3D Forums I really believe this dress is AI added to the little girl. That dress doesn't seem to exist in Daz Studio product land.
*******************************************************************************
I have had Chat GPT to try to find an outfit. I have trained her to look at Daz3d, Renderosity, Poser Content...etc. I get embarrased asking you Daz3d folks about an outfit or hair, etc.
I asked her to find a really cute hair product:
Thought for 33s
I found the hair. It appears to be dForce Bixie Cut Hair for Genesis 9.
Best match: dForce Bixie Cut Hair for Genesis 9 at Daz 3D. It is the short bob/pixie “bixie” style shown with Lyna 9, and the product description mentions the bob/pixie blend, soft face-framing layers, optional sleeker pixie/undercut morphs, and two-tone ash blonde/silver/dark-root materials.
It is also included in the Lyna 9 HD Character Bundle at Daz, along with Lyna 9, Dark Web Dive Outfit, M3D Xyber Vision, and the hair.
PoserContent also lists it as dForce Bixie Cut Hair for Genesis 9, but its “where to buy” entry points back to Daz 3D.
I did not find the exact same hair for sale at Renderosity. The closest Renderosity direction would be searching for short Genesis 9 pixie/bob hair, but the exact match is the Daz product above.
**************************************************************
Turns out, I HAD that hair. I love it!!!
..I had the wonderful opportunity to have met and speak (actually listen more than talk) with Mr Sturgeon at a small Sci-Fi con in Bellingham WA back in 1983 a couple years before he passed on,
That said
I have a background in traditional art that stretches a lifetime so I tend to think visually rather than with words. Hence I am terrible at composing prompts but have little issue "blocking" (a term used in theatre) a scene. Working on a scene in Daz is actually not much different than designing/building a set for stage or cinema and determining where the actors are to go (which is something I've also done.)
Granted with painting and drawing I didn't have to worry about, things like light temperature, lumens, SSS, refraction index and such.
What a thrill to listen to Mr. Sturgeon. A time that will always be with you.
You are one of the artists that facinates me. The animation is brilliant!! Please share how you did that!! It looks like Disney or Pixar...
You don't have much in the Daz gallery. Do you have more somewhere else?
Do you share your story? You have that God given talent that I have wanted forever!!
...thank you very much.
The animation isn't mine, it is a GIF I found many years ago (really love that film and glad they didn't ruin it with a bunch of sequels) don't do much with animation as I have an ancient system which just doesn't have the "horsepower". to handle something like that. I even hold my breath while running a dForce sim, as that sometimes crashes the display driver.
I do have a more comprehensive gallery over on DA I use the same name there. It also includes some very old drawings and sketches I did, the latter when arthritis had set in which eventually made it too difficult to continue.
I use AI with renders because I hate waiting. I render my scene in the pose I want and ComfyUI and Qwen Image Edit {Locally} lets me relight and change my scene however I want. I don't trust AI to place people in the position I want them in and the truth is OpenGL or Filament is enough to get me the renders I want. I create the content, AI relights and changes the lighting and skin and clothes and I can add thing and remove things with no problem.
This is not a correct example.
Street Photographers don't create streets and still it can be art.
The difference is, that you can reinvent a street view and a Stonemason scene view and give it a new meaning.
Generative AI on the other hand presents a scene to you including camera angle, tone mapping ect..., the user's job is to just make selections.
With generative AI, I'd be missing the joy of wandering my camera through a 3d space in order to find that one perfect view to render,
finding that one perfect light set, that somehow resonaces with me and having that one pose and expression on my character, that tells a story.
No generative AI can do this for me. As an artist, I want to project my vision, not just make a selection on a machienes offer of results.
So, declaring generative AI scenes equal to already existing assets or real life scene, shows a lack of understanding the artistic process.
The difference is also, that humans process into art, based on human art work. That matters.
I don't have to respect, what a web trained machien makes of already existing human art.
At the end of the process, there is a human bragging with a result, created by a machiene.
In many ways, AI is like a skilled friend, that can be asked to create some artwork.
Y: "Hey Vincent, make me a nice sun flower field."
V: "Here you are, I made five different versions for you, which one, would you like to keep?"
Y: "I like nr3, but can you add a tree, please?"
V: "sure"
Y: "perfect"
Y: "Hey look, I painted a georgous sunflower field, it's art."
This is generative AI, nothing more.
Again condesing the thought:
I want to know, what a human does with the available resources, not AI.
BTW, I find Stonemason sets frustrating, because they are already so perfect.
The question is always, what to do with those?
There is just nothing left to do potentionally better.
I have a lot of Stonemason products and I almost never use them.
Here is Stonemason's starship bridge rendered in DAZ 3D, and a second image where I had AI convert that image into a comic book look so I could use in a comic I'm creating
The two 3D renders that I am currently working on and editing with AI, each with 30+ layers and layer masks to allow me to paint in what parts of the AI generations I want to keep and manually postwork so that they look as I want them to would disagree.
Those against AI almost always think that all generative AI is "type some words and get a pretty picture". A lot of it is, and that's what we call "slop". Some of it most definitely is not.
I have spent a ton of money on AI digital art to enhance my art. I purchased these as digital scrapbooking kits. This was before I even understood what AI art was.
I have purchased a lot of Dover Clip Art, Design, Pictorial Archive – Dover Publications to use in my art. (Dover is probably all in Public Domain)
So many things I have purchased to use in my art that I didn't have to slave over to produce.
However, I did have to slave over to incorporate these pngs into my own art.
That takes some skill!!!
This is about why and how people use AI. If you don't then this isn't the thread for you, please don't try to turn it into a discussion of AI's putative evils or lack of validity.
Here is my latest work that I started May 29 and completed last night, June 6.
Pretty Little Things by jan_scrapper
I didn't put my work in order, however the renders without bling are mine done in Daz Studio.
I made the bows with Meshy AI after designing them in ChatGPT.
It is a blast to start getting props I need for a particular project.
Items used for Pretty Little Things:
MB Gier for G9F - Daz3d
Addy Florence for G9F - Daz3d
Charlotte for G9F - Daz3d
CHT Memol for G9F - Daz3d
VYK LA Pouf - Renderosity
Elegant Display Props and Scenes - Daz3d
DM's Composed - Renderosity (For lights)
The Swan - FK Design - Renderosity (For the ballerina skirts)
Chat GPT: Used for the decorations of the Pouf Hair Bouquets and the Pouf Bodysuits AND the Title Banner
Meshy AI: To create misc props to use as decorations
I am showing my drafts of renders...a few of many renders.
My PSP workflow had so many layers that I started getting really confused.
You might see many problems...double feet, double hands, etc.
@jan_scrapper: could you describe how you use AI to improve the draping of clothing? Are there limits as to what the AI can do in terms of clothing movement?
Please re read the first post, you missed an important part:
Thanks for posting the before and after. Interesting to see how the AI also improved the bends of the dancers' legs (ie the calf is not clipping into the thigh).
I told Chat GPT:
"isolate woman and her clothes and hat on a separate layer, correct her pants legs and her jacket arms, high resolution, transparent image"
I was still learning Chat GPT. Chat CANNOT do pure transparent png images. So, if I want something that is lace, that is a ton of postwork to get the seethrough item.
Thanks for your reply- that was a very effective correction of those problem areas.
Wasn't that something?? I didn't say much and Chat GPT knew what to do.
It was an on going project, so, after a while, Chat GPT understands a lot of things that I forget to say.
I have told her in the past when I want items isolated, that I want them on a white background and 300 dpi.
Sometimes, I forget that and Chat will go through a monolog of thoughts on the process. An example:
"I believe the user would like to use this in Photo Shop. The goal is a high-resolution that is sutable for printing...."
Things like that. It is so funny when I don't explain myself very well just to see what Chat comes up with :)
This is not a correct example.
Street Photographers don't create streets and still it can be art.
The difference is, that you can reinvent a street view and a Stonemason scene view and give it a new meaning.
Generative AI on the other hand presents a scene to you including camera angle, tone mapping ect..., the user's job is to just make selections.
With generative AI, I'd be missing the joy of wandering my camera through a 3d space in order to find that one perfect view to render,
finding that one perfect light set, that somehow resonaces with me and having that one pose and expression on my character, that tells a story.
No generative AI can do this for me. As an artist, I want to project my vision, not just make a selection on a machienes offer of results.
So, declaring generative AI scenes equal to already existing assets or real life scene, shows a lack of understanding the artistic process.
The difference is also, that humans process into art, based on human art work. That matters.
I have a course coming out in the next few weeks called Art-Official Intelligence. I will walk you through the parallels between using you own artistic expertise and skills- and how they 100% translate from Daz Studio to Daz Studio + AI.
I explain with numerous examples of (classic and established) Photoshop techniques that are 100% the same when using AI.
If I know cameras and lighting in the real world, I will bring that knowledge to Daz Studio. I will use the camera and lighting settings to enhance my Daz renders.
Seems like a simple concept.
I would also do that with AI- use my own knowledge and system, and workflow to translate it for the new tools available. That idea is heavily featured in the course.
I don't need to imagine an "average Daz user" or imagine an "average AI user" to make my arguments make sense. We can all insert ourselves and start from where we are.
Do we all know that you can give AI your camera type/brand, film stock, all associated settings, time of day, position of sun, intensity, and even choose processing for the film? Do we all know you can submit your (personal) photos and artworks for reference? Advanced AI work includes all the same settings that Daz Studio has and more?
Do we know that I can give the AI -images of the village I visited on vacation and use that for the settings or make variations to create a place that looks real, but is not?
It’s Daz Studio + AI for the win. It doesn’t make sense to trade in all the Daz experience, the library and knowledge of environment, lighting, posing etc for the option of typing a few sentences and hoping for the best. It seems more logical to add AI to your toolbox and expand on the art you already make.
I also detail in the workshops how important inspiration is to the creative process. Why am I rendering stuff in the first place? Daz Studio is inspiring. The vendors make stuff that makes me want to use it for my arts. That’s why you buy stuff- from an artist vantage point. Obviously, some of us have commercial concerns and buy the stuff they need for a client/project.
Anyone (and maybe that’s the problem) can prompt their way to nice illustrations.
Most Daz renders are nice illustrations.
Turning the illustration into art is still a very (amazing) human thing. The rules of/to art haven’t changed.
There ain’t no easy way to get to art. That’s the whole point of art. That’s what keeps art, art.
Excellent thoughts here. Thank you for sharing!
I'm curious to find someone to provide a "Comy UI" for the dummy DAZ Studio user. I have decent hardware, but I haven't had much luck getting it to do much to improve my renders. I'm not terribly interested in using the big LLM chat portals. I prefer staying local for everything, which is getting harder and harder to do.
Look up pixaroma on YouTube. That's where I got my ComfyUI start. If you have 16GB VRAM you can run (for example) 14b Wan 2.2 models with minimal swapping.
I'm partial to InvokeAI - I find it to be a more familiar interface for Daz and Adobe users, very image viewport centric. It has layers, masks, local guidance, and you can draw straight into it if you like. It runs locally and has a pretty good model manager for trying out differrnt downloaded models, LoRA's etc. The commercial side of the company was aquired by Adobe about 8 months ago, but it's (free) community supported tool lives on and is still being updated. Their YouTube videos are pretty good and give you a good idea of how the tool can be used.
So did I, when I first learned about the tech three years ago. I spent the rest of the year learning more about it so I could better defend my position, only to wind up adding it to my own toolkit after all. As an artist with a strong, decades-long interest in photorealism, my objections became overshadowed by my need to express myself creatively, so I compartmentalized them and set them aside.
Incidentally, not having to rely on 3D content vendors and their ideas of women's fashion has been a godsend for me.
I want to respond to everyone, however this last sentence really got to me.
I really don't think PA's read the Product Suggestions. Or, they don't read mine, anyway.
I have wanted a Dforce skirt with a ruffle in a circle so I could let it flow over anything I wanted. I still want that.
However, I have used ChatGPT and Meshy AI to get what I want.
I am NOT happy about it, but that is better than NEVER realizing dreams I have had for so many years.
I had to show you guys something!!
I am working on a Broken Doll Repair Shop with a Steampunk theme.
I had Chat GPT make a Steampunk chest with extra arms and legs in it. I have Meshy AI turn these images into props :)
Then, I asked Chat GPT to make a couple of limbs to lay in front of the box.
They thought it might be considered bad.
You just never know :)
Try Grok. I left ChatGPT three months ago when it started rejecting more and more of my prompts - including prompts it had previously enhanced for me. I turned to Grok against my better judgement - I'm not a fan of the CEO - but it's been a helluva lot more useful, and has yet to reject anything I've thrown at it.
I will check out Grok!! I just paid another month of Chat GPT.
I actually re-phrased what I wanted. I guess because I told ChatGPT "to lay in front of this box," a caution was made.
I have the respect that!!!
So, I was able to get what I wanted.
sometimes things such as create a realistic halloween prop of...
or detatched mannequin limbs with red sauce/ketchup spilt and smeared all over can also work