I've been working with poser ever since version 4. I now use Poser 13 and I'm so utterly underwhelmed with Poser 14, that I have no intention of upgrading.

As long as I can remember I wanted to recreate pictures that are in my head. I can draw pretty well, but found it way too complicated/time consuming. So discovering Poser 4 in 2000 was something like a 'godsend'. 

25 years on we're in middle of the AI hype. I'm appalled to see the outpour of mindless AI images drowning out creations that are worthwhile. But at the same time I’m intrigued by its possibilities.

I’ve experimented a lot with AI over the last year. I want to be in control and be able to tinker with the setup, so I started using ComfyUI locally on my laptop. Using this node based open source system was a steep learning curve. And besides learning the technical ins and outs, the most frustrating part is the randomness of AI in going from prompt and latent image to end result.

So my train of thought was: why not try to combine both worlds: use the full control over the scene by creating in Poser and then render and enhance in AI.

After a lot of trail and error I managed to put a ComfyUI workflow together that transforms a very crude Poser render into the picture I have in my head.

This is what the result looks like: 

Being able to transform an original crude Poser render I made into an image that looks almost real and maintains every detail from the original up to the intricate pattern of the fabric, to me was a ‘eureka moment’.

There is still a lot of room for improvement. AI is best at rendering portraits and does a better job at female faces than at male faces (wonder why ;-)). Rendering full body poses or a crowded scene with a lot of props and details in AI, leads to hallucination and loss of detail. If the subject is not looking straight into the camera, eyes tend to go in all directions, except for that in the source image, even with detailed prompting. And often I need to combine 2 or 3 AI renders to create the final image. 

Also I notice that some of my fans on Patreon and other platforms I post on, express resistance to the use of AI.

But in my honest opinion: if Poser (or any other 3D rendering platform for that matter) wants to survive, they need to look into ways to incorporate the power of AI into their products, just like for instance Adobe does in Photoshop.

What do you think?

Cheers,

Richard

Virtualbite.com - Creating the best vampire art you can imagine!

Follow me on Patreon @virtualbite
www.instagram.com/virtualbite_

www.virtualbite.com