Of course, both the previous ideas (http://www.maxwellrender.com/forum/view ... hp?t=28999 & http://www.maxwellrender.com/forum/view ... hp?t=29005) are all very nice but they are soon to become purely academic as there is a new version of XSI (7) out in the very near future, so...

Building Compounds in XSI 7 with ICE

If you haven't already heard (and if not, why not!) the new version of XSI comes with ICE! [read more here - http://www.softimage.com/products/xsi/ice/default.aspx and definitely watch the videos, especially the one called 'Introduction to ICE Deform in XSI 7']

This allows you to compress a huge Render Tree (or ICE Tree) into one node, called a Compound, and you can choose which variables are 'exposed'. This means that if the previous two methods of porting XSI materials (procedural or not) to Maxwell work, then a library of standard materials can be built only exposing certain parameters.

For examples a 'metal' render tree could be built in XSI (which would be converted to an mxm at render time, as described earlier) and only the colour, roughness and nd would be exposed. This would allow you to quickly pull in a 'metal' Compound node, change it's colour and render. Of course you would still have access to all the parameters inside the Compound if you wished to add displacement maps, etc. but this would greatly simplify the creation of many subtly different materials once the cores ones had been built.

Again, thanks for listening. Let's hope the whole process is as easy as I've imagined it (but somehow, I doubt it!).

It looks more like a smooth normal problem. Like i[…]

I had some very strong highlights in some material[…]


Thanks everyone! I didn't want to use volumetrics […]

opening V5 files in v6

just installed latest version for Rhino V6 when I[…]