AmbientShade posted at 2:36PM Fri, 10 November 2017 - #4317570

moogal posted at 1:54AM Wed, 08 November 2017 - #4317562

If Poser had a viewport like blender's Eevee I guarantee more people would start animating.

Not necessarily. A real time renderer would be great in Poser, but one factor I think you might be missing there is poly count. How well does Eevee handle 1million+ poly count scenes? Poser content artists tend to often ignore this aspect of content creation, but there's a ton of content out there that has tens of thousands of polys, sometimes hundreds of thousands, and that sort of content doesn't function too well in a real time engine. In fact it can bring your system to its knees if you don't have a high end machine. I've seen hair models with upwards of 300K polys. Once you dress a single figure in Poser, with hair, shoes, clothes, accessories, you could quite easily be approaching the 500K - 1million poly count, and that's without adding in additional clothed figures, buildings and other props. The reason real time rendering works so well in game engines is because most game engines use very low poly models. Poser generally does not. You also don't see weight mapped joints in game engines because that again draws on processor resources that are needed for the rendering engine to do its job.

Not saying it's impossible, or shouldn't be a requested feature, but content artists would have to start learning how to build models more efficiently, and users would have to start learning how to decimate their models - once the decimation tools that currently exist get refined.

Oh, I've thought about the poly count problem. I don't know if it's the big deal people make it out to be. We're not talking about realtime response, simply using the GPU and modern PBR shaders that work across a far wider range of tools than Firefly or even Cycles materials. Older characters and simple scenes should update in very close to real time, but complex scenes would still display in seconds vs. hours. Again, why limit the development of the program to older content? Part of the reason we have high poly counts is that we didn't have subdivision back in the early days. Now we do, and it would make more sense moving forward to follow iClone's lead of using lighter character meshes and using displacements and subdivision at render time to provide the expected detail. Such characters could be promoted as background figures, as extras, and SM could communicate to users that render speed is dependent on scene complexity to allay unreasonable expectations.
I've long understood that Poser is catering to the still framers, and a lot of stubborn people who haven't kept pace with what is happening elsewhere. When I got into Poser things like Source Filmmaker, Unreal Engine, and iClone didn't exist. But those programs are actually trying to be the program I wanted Poser to be. Ideally I would be mixing Poser figures (stock and custom), iClone content, SFM models, figures from Mixamo etc. in the same scenes if that were actually possible. Perhaps that's why I don't see the need for a new end-all-be-all base figure, as there are hundreds of figures out there I would be using in Poser if the process of doing so were more refined. And as many of those figures were designed to work in a game engine, the poly count thing you mentioned wouldn't necessarily be an issue for me.