dburdick opened this issue on Sep 17, 2005 ยท 13 posts
dburdick posted Sat, 17 September 2005 at 3:15 PM
dburdick posted Sat, 17 September 2005 at 3:19 PM
Djeser posted Sat, 17 September 2005 at 3:52 PM
Whoa, that's excellent, Dave! Thanks again for some great information.
jc posted Sat, 17 September 2005 at 5:35 PM
Yes, thanks - valuable methods!
Veritas777 posted Sat, 17 September 2005 at 6:24 PM
They are really nice looking- but the main drawback of Volumetric Clouds and Fuzzy Materials is that they are awefully slow to render. I've tried Fuzzy Cloud Balls in other renders and while the previews look nice- when it comes render time they are incredibly S L O W. I use an Athlon 64 with 256MB nVidia card and 2 GB RAM and even then I usually bale-out of a render after I see how long its going to take with these kinds of materials.
dburdick posted Sat, 17 September 2005 at 10:34 PM
Veritas777 posted Sat, 17 September 2005 at 11:15 PM
Your scene has a nice foggy kind of effect but the clouds are not really the Fuzzy Cloud Balls that I was referring to. Also- there is very little really "volumetric" happening in your particular scene. Rendering against a relatively plain flat background IS pretty fast- but that's not what I was thinking of... However- if you WANT to use Fuzzy Materials- like Furry-Fuzzy Cloud Balls- I would suggest rendering them against a BLACK PANEL- saving them as 2D Bill Boards- and using the "Flight Simulator Clouds" approach. The 2D versions, with Alpha Channel Masks- actually work quite well. Plus you can multiply them in groups and have the "Fuzzy Volumetric Cloud Look" at hugely reduced CPU rendering time.
jwhitham posted Sun, 18 September 2005 at 6:38 PM
dburdick, I've taken one of the cloud mats out of your scene and blended an angle of incidence node to the density, then applied it to metablob groups for this image. This is getting close to what I'd like, and at last... shadows!
jwhitham posted Sun, 18 September 2005 at 7:18 PM
war2 posted Mon, 19 September 2005 at 2:38 PM
thats looking real good jwithman
jwhitham posted Mon, 19 September 2005 at 6:30 PM
This view from above is interesting; five identical metablob groups at different scalings, look how the functions scale. Also if you make large changes between the scale of spheres within a blob, you get a right mess.
Message edited on: 09/19/2005 18:33
jwhitham posted Tue, 20 September 2005 at 7:45 PM
Djeser posted Wed, 21 September 2005 at 3:54 AM
This is so cool; I spent some time yesterday trying to replicate this with Dave's screenshot printed out, but I'm afraid I need a version for function dummies...