Quote - 3DNeo, I think I understand what you are getting at and I will try to explain the low poly / V4 issue as best as I can in my inebriated state.

Sometimes throwing more brute force processing power won't solve the problem as effectively as streamlining the entire process. Said another way, sometimes less is more. Enter subdivision surfaces.

V4 has a slightly lower poly count than V3 - 68,000 polys verses 72,000 polys respectively or something like that. That many vertices are required for a polygon renderer to create "smooth" silouetted (sp) images (though if you viewed an alpha channel zoomed close enough you would still see facets). Several of these characters will kill your viewport's response and interactivity.

By contrast, take a 3000 poly character and subdivide it. Now you have a smoothly rendered, low weight character that refreshes very quickly in the viewport. This is the industry standard for feature films, and this is the high-end resource of which you speak, with local details provided by sophisticated displacement mapping.

When you are rigging using weightmaps, it is much easier to skin a 3,000 poly character and "wrap deform" it to a subdivision surface than to weight map a 70,000 poly character. There is no debate on this point.

Speaking with the utmost respect and humility, I think your heart is in the right place but your method is not fully realized.

Thanks for your post, it is interesting to see your take on this and get your theories. I can see what you are saying in terms of how other software and such can do it to a better extent. However, can you tell me how Poser can do this WiTHOUT re-writing the entire program? As pointed out, Poser is poly based and thus unless you start over I don't see how the methods you mention are going to be implemented, at least to such an extent as what I was getting at in my prior posts. If you can maybe expand upon this it would nice to talk about this in greater detail, but I don't see how it can be done.

This was basically my point. Unless Poser starts over, which I doubt given the absolute mass amount of add-ons and Daz figures out there, then the only way I see of achieving a HIGH-END detailed renders with photo-realism is to split the software and figures into two distinct markets. Now, if you are talking starting over then yes, everything you say can be done with better means that are not so old. But, current technology with multiple CPUs and cores with HUGE amounts of memory from 8-30GB is no longer a holding point for doing this.

Again, thanks for your thoughts and take on this, I hope that I have maybe explained more of my talking points better with your post as a guide.

Jeff

Development on: Mac Pro 2008, Duel-Boot OS - Snow Leopard 10.6.6 & Windows 7 Ultimate 64bit, 2 x 2.8 GHz Quad-Core Intel Xeon , 10GB 800 MHz DDR2 RAM, NVIDIA GeForce 8800GT.