Well, about the render times, trust me, are faster than the fake volumetrics with a bunch of paralell planes or boxes or wharever to make looks like volumetrics.
I hav not here the exact numbers (Im not connected from my home where I develop everything) but that depend on your machine.
Next week I promise posting about that (Remember that I only get connected once in a week, I’ll be online now for 2 or 3 hours more, so shoot me now!, I dayly check my mail)
But for not waiting so long make some guessings:
All the images have to be shrinked twice to upload
the video have to be shrinked enourmously
I remember that the render that took longer where about 3-4 minutes at 1024-768 in my machine
I’m running on a low-end (the only I could afford) Pentium-D 3.40GHz, 1Gb Ram, Ati Radeon X1550 512 Mb (So no core duo goodies)
Also note that there’s a lot of optimising stuff I will done after everything gets completed at , lets say , 80%
And of course, with the release of testing builds to the comunity bugs,suggestions,improvements will rain, so when I get a solid framework ill make testing builds (Im new to the bazaar)
Note also that the shader implemented is rather simple, so render times could grow depending of what shader ,composite ops you use (I’ll give a lot of options to choose)
Try volumetrics in Ligthwave! they are amazing but also sloooow, in Maya , good volumetrics are also slow, in Pov-Ray, interior media are slow to, everywhere I check theres a fisical law dictating that raycaster could’nt get faster than some point, I’ll try to reach that point.
Also, volumetrics from the stack of textures are faster than from the shader nodes (logical), so for normal animators I would recommend using it rather than the SHD tree.