I've read several of the forums here on Radiosity Island, and I've learned quite a bit in the last few days. I understand that Yves is the undeclared God of radiosity, and I also know that Yves is a veggie cuisine. Though he reminds me most of my High School Programming teacher who worked on flight patterns for Intercontinental Ballistic Missiles during the Cold War using assembly. You've got quite a history. How did you ever end up in animation master?

Anyway... I've decided that I want to do my own project so I'm going to build the Parthenon. And for those of you who aren't familiar with classical greek architecture, it's the image at the bottom of this post. I want to get it to that detailed, and I'm quite certain that it will take a long time to do so. If you guys know any nifty tricks that can speed up the render but not dramatically decrease the quality of this when it's rendered, please let me know.

I've got a column built already, but it's well over 2000 patches in itself. I want to build a displacement texture that will get the general form down correctly, as well as add a normal map for added detail as well as too scatter light. Does anyone here know if there is a normal map creator in hash? I know that I can always turn on fog and make the start of the fog at the very nearest point of the object, and the end of the fog at the back of the object, and set ambience to full and render that for a displacement map. However, there is usually an 8% error in this process of converting it into a normal map with Adobe Photoshop plug-ins; one of the reasons why I wanted to know if Hash supported that feature. The other issue is blending them together. If and when I find a way to obtain a normal map the question is, is there away to mix them with the displacement map to make the object appear more realistic?

Post Statement
{
Off the topic a little, does your guy's versions of animation master crash a lot?
}
End Post



The Parthenon

I agree. 2000 patches for a single column is way too many. You need to show us a wireframe. I get the impression that you are approaching modeling with patches like modeling with polygons.

Sorry my bro isn't here right now... but here's a picture of one of his columns.

By the way... how do you get a programmers lable?

Do you need to model the segment cracks? Will this be viewed from a close-up? If not, then use a bump or normal map. Also you could get the same ribbed form with half the number of vertical splines by adjusting the biases.

I got a "programmer" label because I am (among other qualities. I actually should have 3 stars for contest won but I like to keep a low profile. ). A:M users who have programmed plugins for A:M are labeled "programmer" too.

The only thing is that I need the inlays for the columns. There are twenty inlays so that acounts for 40 sides to the cylender. If I were to cut down on the number of sides, it would get ride of that detail. Also, the columns are stacked ontop of each other, so I wanted to achieve that effect when rendered. It will also be easy to edit each stacked column part. The Cieling Cap is modled with only 10 sides, and the block on the top is beveled with... 6+4+4+4 = 16 sides or 18? I only really know how I could cut off a few parts of the columns. If all of the columns were to stay in the same patch count, it would be approaching 100,000 patches. That's not even the inner columns, the outside rocks, the roof, the steps, or even the ground. So I'm really thinking that maps are the way to go.

Change the bias? you mean the magnitue handles? BUT LOOK AT THE DETAIL MAN!! it's somewhat breath taking... espacially when it's all together

Did you turn on oversampeling to show that? Or even antialiasing? Because it looks almost like a patch/gamma related mess up, and I know that I made it so it wouldn't do that.

More like a print screen. Dont cha know

If you are going for Normal Maps, these threads may be useful:

• Normal maps and v12?
• Normal Map Experiments

I think you could lower the patch count significantly and still maintain modeled seams by re-thinking their spline profile. My suggestion is shown in the attached image.

I think I can use maps for the cracks, but I'm probably still going to need the 40 sided cylender because of the 20 inlays. Every other point is depressed into the cylender and the other points are set to a magnitude of 20 for rounded edges.

Corenthian columns would be out of place with the Parthenon. It was built far before that style came around.

Just to let you guys know, I haven't stopped working on the Parthenon. The acropolis itself is a wonder, so I'd like to see the whole thing added. Maybe I could do both a current view and an old view of it. I think I'm going to go with normal maps for the medapies, trygliphs, and the freeze. Detail is great, but boy does it add to render time (A network A:M Cray Supercomputer isn't sounding too bad now is it *wink* *wink*).

ehh.. remember that the columns curve outward a bit right? the steps are slightly concave as well.

Oh, I'm reduing my models. Their going to be much faster to render now. I'm going to go with high quality normal mapping and textures, instead of high patch geometry.

And yes, the Doric columns are smaller at the top than at the bottom. 85% to 80% their base diameter. The Greeks were acurate, they weren't perfect.

I mean they don't taper in a straight line. They bulge out a bit in the middle.

The Greeks knew what they were doing. Those tapering columns were designed that way for forced perspective:

http://www.math.nus.edu.sg/aslaksen/projec...e/parthenon.htm

Yep, the Greeks did. The same was done for this well known place at Disney Land.