Very cool! I stand corrected. :)
On Tue, May 28, 2013 at 1:15 PM, Aghiles Kheffache <aghil...@gmail.com>wrote: > Hello Ciaran, > > Next release of 3Delight for Sotimage is a path-tracer. REYES is optional. > (You can select one or the other). > > The path-tracer introduces some novel sampling techniques, such as nice > sampling across multiple bounces, high quality environment sampling (no > need to blur your maps) and of course fast render region re-rending (for > now only light, camera and material edits), … > > Regarding performance : some of you will be surprised. ;) > > -- > Aghiles > www.3delight.com > > > > On 2013-05-28, at 12:33 PM, Ciaran Moloney <moloney.cia...@gmail.com> > wrote: > > True, it originated as a REYES renderer, but now it may be better thought > of as a hybrid renderer taking advantage of both techniques when needed. > > I like working with Arnold too, but I don't think fast renders are one of > its strongest selling points. > > Ciaran > > On Tue, May 28, 2013 at 4:48 PM, Alan Fregtman <alan.fregt...@gmail.com>wrote: > >> I haven't tried 3Delight, but Arnold is **way** too good. My gut feeling >> is Arnie is probably faster overall. >> >> That said, they're entirely different rendering architectures so it's not >> fair to compare. 3Delight is (as I understand it) renderman-compliant which >> implies they probably use the Reyes rendering model: >> http://en.wikipedia.org/wiki/Reyes_rendering >> whereas Arnold is not a reyes renderer, but a superoptimized "hardcore" >> raytracer: >> http://en.wikipedia.org/wiki/Ray_tracing_(graphics) >> >> It would be more fair to compare 3Delight against Pixar's Renderman, and >> Arnold against mentalray, than Arnold to 3Delight or Renderman to mentalray. >> >> Just my $0.02, :p >> >> -- Alan >> >> >> >> On Tue, May 28, 2013 at 11:28 AM, Leonard Koch >> <leonardkoch...@gmail.com>wrote: >> >>> That entirely depends on the use-case. >>> Arnold is in faster most raycasting situations, but 3Delight will be >>> faster for for example motion graphics or character animations with >>> unrealistic shading. >>> And even then it can very much depend on the complexity of the scene >>> and on of what kind that complexity is. >>> Then there is also the aspect of artist time spent to get to that >>> render-time, which - depending on how easy it is to get from the default >>> state to the final render state - can wildly differ in between the >>> different renderers and shots. >>> The speed of a renderer in a production is a very complex metric and >>> can't really be summarized into a simple statement like "Arnold is faster >>> than 3Delight". >>> >>> >>> On Tue, May 28, 2013 at 4:19 PM, Christopher < >>> christop...@thecreativesheep.ca> wrote: >>> >>>> Is it as fast as Arnold :) ? >>>> >>>> >>>> Morten Bartholdy wrote: >>>> >>>> Looks interesting Greg. Two questions: >>>> 1. What is your hardware setup ? >>>> 2. Could you upload it in higher resolution so we can read the menus? >>>> >>>> MB >>>> >>>> >>>> >>>> Den 28. maj 2013 kl. 15:37 skrev Gregory Ducatel >>>> <gduca...@gmail.com><gduca...@gmail.com>: >>>> >>>> >>>> Hi Guys, >>>> >>>> In case you did not had a chance to look into the next version of 3dfs, >>>> here is a little test I did. >>>> >>>> http://vimeo.com/66105781 >>>> >>>> Cheers, >>>> >>>> Greg >>>> >>>> >>>> >>>> >>>> >>> >> > >
