hi guys, i am impressed that this topic gets quite a lot of feedback! thx for all your contributions! For me it would be totally ok if we end up flat after the deepVectorBlur.
My usecase is just day-to-day 2d-post-motionblur which everyone of us does with the vectorBlur node i think....this does the job in most of the case where you don't have rotation or nonlinear motion in subframes.... The only thing that is super anoying is that this workflow always stops working on overlapping objectborders.....i don't care about what may be occluded, but appling the motionvector on every deep sample will definitely help getting rid of 99% of all artifacts i guess. @colin: this is also where you guys come into play....since i know that bokeh does a pretty good job at defocusing deep images i guess it's you that need to incorporate the vectorBlur directly inside of bokeh, because no matter how - you need to do defocus and vectorBlur at the same time i guess or one of the informations will be lost....do you think this is doable? @all: do you agree on the idea of doing defocus/zBlur and vector/motionBlur in one node? cheers Oli @patrick: all is good over here! cheers from munich to LA! 2013/12/11 Michael Garrett <[email protected]> > Just wanted to say, I see having a DeepVectorBlur as a potentially great > way of minimising deep image file size since there would be far less deep > samples stored. I'm finding that even for fully opaque non-volumetric > surfaces that I know will never intersect with anything, there is "sample > bloat" with a lot of renderers. The trade-off with storing the vector is a > somewhat increased file size but that would be greatly offset by the vastly > reduced number of deep samples. > > Doing this with a flat output seems like a real option, like pgBokeh. In > fact, one of the Foundry devs mentioned he'd like to see this, since motion > vectors were revamped and we now have deep output from the ScanlineRender > node. In fact it already outputs deep motion vectors. > > Michael > > > On 11 December 2013 12:52, Elias Ericsson Rydberg < > [email protected]> wrote: > >> If the goal is to visualise motion blur using deep points then the >> difficulty goes up quite a bit. If The goal is to store dV/dT for every P >> then it becomes a lot easier. What would be the benefit of having the blur >> visible in 3D space? It's gonna end up flat in the end anyways? >> >> Cheers, >> Elias >> >> 11 dec 2013 kl. 18:34 skrev [email protected]: >> >> > Also; blurring an element would reveal areas behind it, for which there >> are no samples. Same as in the non-deep case. >> > >> > eetu. >> > >> > >> > On 2013-12-11 18:55, Steve Newbold wrote: >> >> I don't think it would destroy deep data, but for every sample you'd >> >> have X number of additional samples all with a lower alpha value but >> >> accurate deep data. It the creation and processing of these additional >> >> values that make this really not feasible in reality, or at least with >> >> current equipment. >> >> >> >> Steve >> >> >> >> On 11/12/13 16:49, Neil Rognvaldr Scholes wrote: >> >> >> >>> Hi guys im intrigued - how would you sue this deep vector blur? >> >>> >> >>> To my mind , blurring deep data would actually destroy acurate deep >> >>> values - no? >> >>> >> >>> Like blurring P ? >> >>> >> >>> Curious. >> >>> >> >>> Neil Rognvaldr Scholes >> >>> >> >>> +44 (0) 7977 456 197 >> >>> www.uvfilms.co.uk >> >>> On 11/12/13 06:55, Patrick Heinen wrote: >> >>> >> >>>> Hi Oli, >> >>>> >> >>>> I would also be very interested in knowing if someone has that >> >>>> working already. I was thinking about writing such a plugin a year >> >>>> ago but never actually did du to the lack of c++ knowledge, time >> >>>> and a renderer that spits out deep vector pathes. Do you have a >> >>>> renderer that puts out deep vector pathes? I would take an >> >>>> educated guess and say it shouldn't be all to hard to write such a >> >>>> plugin, but there's probably some little detail I miss ;) >> >>>> big +1 for more deep tools though! >> >>>> >> >>>> Hope everything is good in Munich! >> >>>> cheers, >> >>>> Patrick >> >>>> >> >>>> On 09.12.2013, at 15:37, Oliver Markowski <[email protected] >> >>>> [4]> wrote: >> >>>> >> >>>>> Hi guys, >> >>>>> >> >>>>> i hope this hasn't been posted before, but a quick search did >> >>>>> not show anything... >> >>>>> >> >>>>> Does anyone already have a solution for proper deepVectorBlur? >> >>>>> Would it be easy to do as a simple-plugin or is this more >> >>>>> sophisticated? >> >>>>> >> >>>>> Maybe someone from thefoundry could also tell me if this is >> >>>>> already on their todo list... >> >>>>> >> >>>>> cheers >> >>>>> Oliver _______________________________________________ >> >>>>> Nuke-users mailing list >> >>>>> [email protected] [1], >> >>>>> http://forums.thefoundry.co.uk/ [2] >> >>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >>>>> [3] >> >>>> >> >>>> _______________________________________________ >> >>>> Nuke-users mailing list >> >>>> [email protected], >> >>>> http://forums.thefoundry.co.uk/ >> >>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >>> >> >>> _______________________________________________ >> >>> Nuke-users mailing list >> >>> [email protected], http://forums.thefoundry.co.uk/ >> >>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> >> >> >> >> >> >> Links: >> >> ------ >> >> [1] mailto:[email protected] >> >> [2] http://forums.thefoundry.co.uk/ >> >> [3] >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> [4] mailto:[email protected] >> > >> > _______________________________________________ >> > Nuke-users mailing list >> > [email protected], http://forums.thefoundry.co.uk/ >> > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> _______________________________________________ >> Nuke-users mailing list >> [email protected], http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> > > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >
_______________________________________________ Nuke-users mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
