Hi guys - Sorry you are finding it frustrating. This is definitely not all the way to where we want it to be, but most complicated features come down to a choice of break it up into multiple steps and release the improvements each step of the way, or wait until everything is done to release any of it. Basically, we have finite resources and it's going to take the time it takes to do the work so you can have some of it along the way, or nothing until it's all done. I'm of the opinion that it's better to have something as soon as possible but there is always the danger that we get disappointment for not having everything there perfect all at once.
This release gets the splinewarper onto the same spline library and UI interaction as the roto so the UI is consistent and you have access to the individual points via the curve editor and dope sheet and can set expressions on them, copy between the nodes, etc. but some of the underlying limitations of the previous version remain. For the next release we are planning work on the spline system internals that will allow us to do completely independent splines for the source/destination as well as overall performance improvements which will affect paint, roto, and the spline warper now that they are all on the same base. There are a couple other things that I think are still painfully missing, specifically, one, being able to edit the source or dest shape through a partial warp, so for instance you have it set to 75% warp from source to dest and you can see and edit the source or dest spline "through" that view of the shape and, two, to be able to render a mask for each shape and cut it out of the background in the spline warp node. That's all in the big picture; it just comes down to balancing resources and competing priorities to try to get the most out of each iteration. As always, constructive criticism, specific use cases, and requirements like you described are greatly appreciated. It helps us understand the big picture we are aiming for and when we prioritise, what's most important amongst the details. Matt -- Matt Plec - Senior Product Designer - The Foundry 6th Floor, The Communications Building, 48 Leicester Square, London, WC2H 7LT, UK Tel: +44 (0)20 7968 6828 • Fax: +44 (0)20 7930 8906 • Web: www.thefoundry.co.uk The Foundry Visionmongers Ltd • Registered in England and Wales No: 4642027 On 2 Aug 2011, at 18:24, Ean Carr wrote: > I was just playing with that today and had all the same thoughts as you. -E > > On Tue, Aug 2, 2011 at 6:22 PM, Bill Gilman <[email protected]> wrote: > hear hear > > On Aug 2, 2011, at 10:16 AM, J Bills wrote: > > > just trying it for the first time. having a hard time trying to get it to > > do what I want and several things seem counterintuitive. anyone else out > > there having success with it? am I missing something? > > > > I've got gobs of roto for this shot I'm on, and all I want to do is warp a > > clean plate for a few frames to remove a rig. simple enough - seems like a > > textbook case. > > > > should be as easy as: > > -copy roto already done of the area I want into splinewarp node > > -delete source keyframes on roto except for key on clean plate frame > > (reference frame) - so if clean plate was painted on frame 45 I'd delete > > all keys except for 45 on the source. > > -leave destination keys alone so that the source warps from frame 45 > > position to current frame using original roto > > -then I just play with correspondance points, pinning, warp settings and > > boundaries until it looks good. > > > > nope, can't seem to get the roto in there and working. > > > > I can copy and paste it in but beyond that the shapes are very hard to work > > with. instead of independant source and destination shapes, they appear to > > be the same shape! I can't seem to get access to one set of keys without > > manipulating the other. If I try to delete out all of the keyframes on my > > 'source' shape, it just deletes the destination keys as well. > > > > I thought the whole point of redesigning it was to make it play nice with > > the new roto tools and make everything interchangable like it was in shake > > (and ER before it), where you could easily shuffle roto in and out. > > imperative for good warping, because eventually those same shapes become > > your matte to help blend the warped area with the original plate. > > > > as it is, unless I'm missing something (which I hope I am!) it seems like > > we're expected to originate all of the roto within the node and are unable > > to edit it much after the fact independently between source/destiantion. > > I'm not quite sure what that type of workflow is good for, but apparently > > it will warp the face of a lion from the manual nicely. will have to > > remember that next time that comes up. :-/ > > > > > > _______________________________________________ > > 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
