Thanks alot, that solved the problem. :) On Sun, Feb 27, 2011 at 12:12 AM, Michael Garrett <michaeld...@gmail.com>wrote:
> I've been using an ImagePlane-type of approach for roto tasks a lot > recently, but have supplemented it with finding the Z distance between > two 3d points (camera and subject) to improve its accuracy. I could be > misinterpreting what you mean by "depth rotation" but I think at the > ref frame you need to set the roto card distance from camera to be as > accurate as possible then it will stick as the camera rotates round > the object. If it's inaccurate (too close) that would explain why the > card is zipping off to the side of screen. > > So if you have a PFTrack track with null points imported into Nuke as > Axis nodes then you need to find the distance between an Axis located > where the roto subject is, and the camera position on the ref frame. > > http://freespace.virgin.net/hugo.elias/routines/r_dist.htm > (shows how to calculate the distance between two points, plug that > into an expression node then pass it to ImagePlane's distance value). > > > Michael > > > On 24 February 2011 04:57, Jon W <jwesst...@gmail.com> wrote: > > Thanks for your tips, I have however just bit the bullet and made the > roto > > by hand. > > > > Frank: That gizmo was quite cool, definitely something i will use in the > > future. Only one problem, it wants to compensate for the cameras > > depth-rotation (unsure what the proper name is) which causes the image to > be > > off screen for the majority of the shot, you can somewhat compensate for > > that by changing the refFrame but not entirely. > > > > Do you know if there is any way to disable that rotation for the gizmo? > > > > Thanks. > > On Tue, Feb 22, 2011 at 11:45 PM, Frank Rueter <fr...@beingfrank.info> > > wrote: > >> > >> maybe the ImagePlane gizmo helps? It's kinda like the Tracker node for > 3D > >> in that it projects the image (in your case the roto) onto a card at a > >> reference frame, then uses the camera data to move it from there on. > >> http://www.nukepedia.com/gizmo-downloads/transform/imageplane/ > >> > >> > >> > >> On Feb 23, 2011, at 4:20 AM, Shailendra Pandey wrote: > >> > >> > you would use a reconcile 3d node to generate 2d tracking data from > >> > the matchmove which can be fed into a 2d tracker node to generate > >> > stabilize and track data to do the roto. > >> > Hope it helps > >> > > >> > Shailendra > >> > > >> > > >> > > >> > On 2/22/11, Jon W <jwesst...@gmail.com> wrote: > >> >> I tracked a 325 frame long shot in PFTrack (to use in Maya), but I > also > >> >> have > >> >> to do some rotowork, and the camera is pretty jittery so I figured it > >> >> would > >> >> perhaps be easier to use my PFTrack to make it 100% stable, do my > roto > >> >> and > >> >> then reverse the stabilization. > >> >> The shot rotates 180° and moves backwards (sort of like a helicopter > >> >> circulating a building), thus I would not want to use any > >> >> Z(?)-rotation. > >> >> > >> >> Thanks > >> >> Jon > >> >> > >> > _______________________________________________ > >> > Nuke-users mailing list > >> > Nuke-users@support.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://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users > > > > > > _______________________________________________ > > Nuke-users mailing list > > Nuke-users@support.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://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >
_______________________________________________ Nuke-users mailing list Nuke-users@support.thefoundry.co.uk http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
