So I was right. Accept for overhead part (kidding joking, smart ass alert)
I just remember a lot of people asking in like V5(or 6?) when the first new paint node was added that a clean roto node exist, and as some point it was added.) Randy S. Little http://www.rslittle.com/ http://www.imdb.com/name/nm2325729/ On Wed, Apr 1, 2015 at 6:55 AM, Frank Harrison <fr...@thefoundry.co.uk> wrote: > Hey, > > just because i am too lazy to switch the viewer to display the alpha >> channel >> > Did you know that when you press the "A" key in the viewer it allows you > to switch between Alpha and RGB? :) > > On a side node: If i, lets say use a blur node set to use rgba, but >> downstream only use alpha from that (i.e. as a mask input on another node), >> will Nuke be slower because it blurs rgba instead of alpha only? Or will >> only the requested channels be calculated upstream? > > It should just operate on the Alpha channel and if the blur node is just > operating on Alpha then that should be all that it operates on and all that > it *directly* supplies to its down stream Nodes. If you're viewing the > Blur with RGB enable in the viewer then you'll pull those channels through > the blur. > > So I guess you might (depending on you graph) be able to work faster if > you operate on a channel at a time. > > When you mention expressions, do you mean expressions on the Roto/Paint >> nodes themselves, or also generally in the script? >> > In this case Roto/Paint specifically. I've seen scripts with expressions > on most of the points on most of the shapes. > > And what would you consider excessive? :) >> > Well, expressions are an incredibly useful feature and help us do lots of > very cool stuff very quickly, but, by their very nature Nuke can not work > out when the value that an expression evaluates to has changed; this means > that Nuke has to evaluate Nodes (and sub-graphs with Nodes) with > expressions more often, which can lead to slowdowns, sometimes. > > I hope that answers your questions :) > > F. > > > On 1 April 2015 at 10:37, Daniel Hartlehnert <dah...@gmx.de> wrote: > >> Oh, i forgot to add that i have only worked with v7 so far, no experience >> with 8 or 9 yet. >> >> >> Am 01.04.2015 um 11:17 schrieb Frank Harrison: >> >> Hey Daniel, >> >> That's fascinating to hear, I wonder if it does relate to the bumber of >> channels they use by default? >> >> There are various things you can do to make Roto and Paint appear slower, >> such as use excessive numbers of expessions or view a node downstream of a >> Roto with lots of comptationally expensive Nodes in between. >> >> If anyone does have any reproducable performance metrics or scripts >> they'd like to share we'd love to see them, add them to our autotests and >> make it a better experiance :) >> >> F. >> >> On Wednesday, April 1, 2015, Daniel Hartlehnert <dah...@gmx.de> wrote: >> >>> Hi Frank, >>> >>> thanks for an "official" answer :) >>> >>> My experience differs though, so i don't have any "metric" to base it >>> on, Marten. Only common sense and experience. Therefore my answer i gave >>> before. >>> >>> >>> Am 01.04.2015 um 10:00 schrieb Frank Harrison: >>> >>> Hey Guys, >>> >>> Under the hood the two Nodes are exactly the same, they just cater to >>> different work flows. >>> >>> As a result, the only time RotoPaint should be slower is when you have >>> paint/clone/smear strokes interleaved with bezier/bspine shapes. >>> >>> F. >>> >>> On Wednesday, April 1, 2015, Daniel Hartlehnert <dah...@gmx.de> wrote: >>> >>>> Isn't it obvious? Roto node has much less functionality, hence it its >>>> much faster to process and has a smaller memory footprint. Bugs from the >>>> paint part of Rotopaint cannot destroy your script, Rotopaints with more >>>> than 100 strokes tend to slow down your script (if i ever see that progress >>>> bar from a rotopaint node i start to ...) >>>> Rotopaints are responsible for most broken scripts in my experience. >>>> >>>> Wyh would you use Rotopaint if you don't actually paint? >>>> >>>> Am 31.03.2015 um 20:14 schrieb Simon Björk: >>>> >>>> For some reason I always reach for the RotoPaint node instead of the >>>> regular Roto node. I'm not sure why I got into this habit, but it might be >>>> (if I'm not mistaken) that the RotoPaint node was introduced before the >>>> Roto node back in 6.0. >>>> >>>> Anyway, does anyone know if there's a difference in >>>> performance/stability when doing just regular roto? We've all had or >>>> problems with the RotoPaint node over the last couple of versions, but I >>>> have never actually compared the two nodes to see if one is "better" than >>>> the other. >>>> >>>> >>>> >>>> >>>> >>>> ------------------------------- >>>> Simon Björk >>>> Compositor/TD >>>> >>>> +46 (0)70-2859503 >>>> www.bjorkvisuals.com >>>> _______________________________________________ >>>> Nuke-users mailing list >>>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>>> >>>> >>>> >>> >>> -- >>> Frank Harrison >>> Senior Nuke Software Engineer >>> The Foundry >>> Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 >>> Web: www.thefoundry.co.uk >>> Email: frank.harri...@thefoundry.co.uk >>> >>> _______________________________________________ >>> Nuke-users mailing list >>> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>> >>> >>> >> >> -- >> Frank Harrison >> Senior Nuke Software Engineer >> The Foundry >> Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 >> Web: www.thefoundry.co.uk >> Email: frank.harri...@thefoundry.co.uk >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> >> >> >> _______________________________________________ >> Nuke-users mailing list >> Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ >> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >> > > > > -- > Frank Harrison > Senior Nuke Software Engineer > The Foundry > Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 > Web: www.thefoundry.co.uk > Email: frank.harri...@thefoundry.co.uk > > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >
_______________________________________________ Nuke-users mailing list Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
