Thanks Peter, I would have seriously begun contemplating your parallel universe theory but the problem seemed to have been the strands. I'm still not sure what causes the strands to be so long to mesh the closer you get to them though.
2013/12/13 <pete...@skynet.be> > something easily overlooked: a close up could fill the whole of the > image, while a long shot might have for example 10% of the image area > covered with the subject. In which case a ten fold difference in rendering > time is totally normal. > > other than that, any combination of stochastic sampling - ambient > occlusion, area lights, soft reflections/refractions, DOF and/or MB ... - > will slow down mental ray massively in close ups – in my experience more so > than you would expect taking into account the image area coverage mentioned > above. > MR’s is sending endless amounts of rays per pixel in recursive loops – I > suspect MR of actually sending samples to a parallel universe where > rendering gets exponentially faster the more samples you ask – but at this > point this is just conjecture. > > adapting sampling settings, using unified sampling or switching to the > rasterizer are all things that might alleviate the pain. a little bit. > > > > *From:* Antonin Messier <antoni...@gmail.com> > *Sent:* Friday, December 13, 2013 5:10 PM > *To:* softimage@listproc.autodesk.com > *Subject:* Re: Closeup frames super long to render. > > Hi, yes, Mental Ray, Vegetation is mostly deformations on curves, there > are ICE strands as well but not in the closeup. > > > > > *Antonin Messier Turcotte*3D and Compositing Artist > Fly Studio > > T: 514-490-1117 > M: 514-743-4211 > E: antoni...@gmail.com > www.flystudio.com > > > 2013/12/13 Eric Mootz <e...@mootzoid.com> > >> Hi Antonin, >> >> What is your vegetation made of? >> Are there any particles involved? >> Are you rendering with Mental Ray? >> >> Cheers, >> Eric >> > >
