DAZ's survival in the AI era hinges on its lighting capabilities

iSeeThis

I've invested a significant amount in DAZ over the years. Frankly, I didn't anticipate the advent of the AI era. Had I known, I might have allocated only 5% of my resources to it.

However, with the widespread adoption of AI, I've come to realize that achieving state-of-the-art visuals still demands considerable effort on my part. Even with AI tools, creating professional-grade artwork requires a combination of DAZ and additional refinement through AI editors. There's no single solution that covers all bases. (It's disheartening when someone casually remarks, 'Oh, the AI did a great job. Your work looks fantastic!' It undermines the hours of creative input I've put in.)

Reflecting on my experiences, I believe that for DAZ to remain relevant in the next decade, enhancing lighting capabilities is crucial. If we can fine-tune everything else and achieve realistic results akin to those produced by AI, then we're onto something significant! This strategy might mean sacrificing some revenue from lighting products initially, but it will ultimately attract more devoted followers and increase overall income.

lilweep

not sure how you reaching your conclusions.

So in your mind AI can do everything but light a scene lol? Wouldnt it also be able to do that?

Daz already has ray tracing.

Daz remaining relevant in the next decade which is a long time, will require more than making ray tracing better whatever that means.

Also, do people use lighting products? light is free.

FSMCDesigns

"widespread adoption of AI"? I want nothing to do with AI, especially here at DAZ. AI is doing the work. I can edit and tweak materials and shaders all i want, the renderer is still doing the work.

plasma_ring

iSeeThis said:

If we can fine-tune everything else and achieve realistic results akin to those produced by AI, then we're onto something significant!

Fortunately, 3D software allows us to place virtual lights in configurations similar to those used in real studios and use other tools to simulate realistic lighting. 

Lighting in generated images that are stylized the way artwork is usually mimics concept art or game engine renders. That lighting tends to be the result of deliberate design decisions made by individuals or teams who know how to make it look good. The AI is not achieving state-of-the-art lighting as a consequence of the technology; it's referencing what we know looks good and what we've already achieved. 

This strategy might mean sacrificing some revenue from lighting products initially, but it will ultimately attract more devoted followers and increase overall income.

Probably worth remembering that the revenue sacrificed by making specific types of products obsolete has a really good chance of being an individual person's income. 

WendyLuvsCatz

well you can use relight on your renders https://clipdrop.co/relight

savagestug

Achieving "state of the art visuals" is inherintly supposed to require considerable effort. If you just type in prompts trolling off other people's work, that is meaningless.

That said, I would be giddy if we had "include" and "exclude" parameters on Iray lights, like in 3DS Max.

rcourtri_789f4b1c6b

iSeeThis said:

I've invested a significant amount in DAZ over the years. Frankly, I didn't anticipate the advent of the AI era. Had I known, I might have allocated only 5% of my resources to it.

However, with the widespread adoption of AI, I've come to realize that achieving state-of-the-art visuals still demands considerable effort on my part. Even with AI tools, creating professional-grade artwork requires a combination of DAZ and additional refinement through AI editors. There's no single solution that covers all bases. (It's disheartening when someone casually remarks, 'Oh, the AI did a great job. Your work looks fantastic!' It undermines the hours of creative input I've put in.)

Reflecting on my experiences, I believe that for DAZ to remain relevant in the next decade, enhancing lighting capabilities is crucial. If we can fine-tune everything else and achieve realistic results akin to those produced by AI, then we're onto something significant! This strategy might mean sacrificing some revenue from lighting products initially, but it will ultimately attract more devoted followers and increase overall income.

While Daz's relevance and business viability in an era of generative AI applications is worth questioning, I'm not sure I would put lighting at the top of the list of areas to improve.  While lighting-related problems (inadequate light level, excessive light level, different shader types with different lighting needs used in the same picture, etc.)  are probably the most common technical failing I see in renders, that's more of a user knowledge and capability problem than a limitation of Daz Studio.  Besides, lighting is heavily tied in with the render engine.  Since Daz hasn't developed an in-house proprietary render engine, they're somewhat at the mercy of nVidia.

Particular lighting products in the store are at risk of becoming obsolete due to changes in either Daz Studio or the available render engines.  That's just the nature of changing technology.  But it is fair to ask if the company is just too stagnant at this point.  Has there been an improvement in animation in the last eight years?  Has there been a significant improvement in clothing fit in the last five years?  Does hair really look better in renders than it did five or six years ago?  Can you make a better render than you could in 2019?  Is it getting faster or easier to make a good render?  Is the company really invested in anything besides a daily release schedule of content that's usually completely fungible with already released content?

fred9803

Just better lighting capabilities in AI? How about far superior realism, human-looking people, realistic and naturally fitting clothing, lookalikes that look like the actual person and renders that take seconds rather than hours. I can see why people are switching away from 3D software.

Edit - forgot to mention.... and everything for free.

PerttiA

rcourtri_789f4b1c6b said:

While Daz's relevance and business viability in an era of generative AI applications is worth questioning,

I don't understand why.

With AI, one gets a picture of something resembling the description one has typed. 
With DS one gets the real 3D object that one can manipulate and take pictures of (renders)

It's the same as owning a Ferrari and taking pictures of it, versus searching for Ferrari pictures in Google.

Richard Haseltine

Closing this thread after having to remove a swathe of posts, mostly for offensive comments (or replies thereto).

Richard Haseltine

wurger said:

Achieving "state of the art visuals" is inherintly supposed to require considerable effort. If you just type in prompts trolling off other people's work, that is meaningless.

That said, I would be giddy if we had "include" and "exclude" parameters on Iray lights, like in 3DS Max.

I am not familiar with 3DS, so this may not be an exact match but Light Path Expressions, a type of Canvas available in the Advanced tab of Render Settings, do allow control of what interacts with which light.