Well, that is weird as it's working fine on my side, but it has been done in 2014. I'm boarding for a 24h flight, I'll have plenty of time to check what's going on. Unfortunatly I don't have 2013 with me... In the worst case I have a scripted version for the init part, it's just less pratical as ICE attributes are locked when in use. Cheers
> Le 18 déc. 2013 à 15:34, Alok Gandhi <[email protected]> a écrit : > > Hi Ahmidou, > > I am getting the same error as Jeremie. From what I concur, the phi > distribution on sphere node furnishes an array of position which is plugged > into the direction of the a raycast node. The direction can be either a > single vector or vector per point of the deformee while you are are supplying > an array of vector, thereby causing incompatible types error. > > I did not have time to compare your implementation with the paper to further > debug this, but this can give you an direction to further solve this issue. > > Either that or we are not setting up in the tree in the correct way. > > >> On Tue, Dec 17, 2013 at 9:09 PM, Jeremie Passerin <[email protected]> >> wrote: >> Not sure, I'll try again tomorrow morning.. >> >> if anyone else on the list is testing it, let me know if it works for you. >> >> thanks again for sharing, that might be a nice addition to my tool set ! >> >> >>> On 17 December 2013 17:32, Ahmidou.xsi <[email protected]> wrote: >>> I'm not in front on my computer, but what kind of error did you got? >>> >>>> Le 18 déc. 2013 à 11:52, Jeremie Passerin <[email protected]> a écrit : >>>> >>> >>>> Nice job Ahmidou ! >>>> Looks pretty straight forward but I wasn't able to make it work yet. :D >>>> I got an error on the phi distribution on Sphere... I might be doing >>>> something wrong. >>>> I'm using XSI 2013, could that be it ? >>>> >>>> Here is what I've done >>>> >>>> >>>> # python >>>> xsi = Application >>>> >>>> # Your path here ! >>>> compoundPath = >>>> "\\\\source\\source\\dev\\passerin\\blurdev\\workgroups\\xsi_blurdev\\data\\Compounds" >>>> >>>> xsi.CreatePrim("Sphere", "MeshSurface", "", "") >>>> xsi.SetValue("sphere.polymsh.geom.subdivu", 24, "") >>>> xsi.SetValue("sphere.polymsh.geom.subdivv", 24, "") >>>> xsi.FreezeObj("", "", "") >>>> xsi.CreatePrim("Cube", "MeshSurface", "", "") >>>> xsi.SetValue("cube.cube.length", 9, "") >>>> xsi.ApplyTopoOp("TriangulatePolygons", "cube", "siUnspecified", >>>> "siPersistentOperation", "") >>>> xsi.FreezeObj("", "", "") >>>> >>>> xsi.SelectObj("sphere", "", True) >>>> xsi.ApplyOp("ICETree", "sphere", "siNode", "", "", 0) >>>> xsi.AddICECompoundNode(compoundPath+"\\initPMVC.xsicompound", >>>> "sphere.polymsh.ICETree") >>>> xsi.ConnectICENodes("sphere.polymsh.ICETree.port1", >>>> "sphere.polymsh.ICETree.initPMVC.Execute") >>>> xsi.SelectObj("cube", "", "") >>>> xsi.AddICENode("GetDataNode", "sphere.polymsh.ICETree") >>>> xsi.SetValue("sphere.polymsh.ICETree.SceneReferenceNode.reference", >>>> "cube", "") >>>> xsi.ConnectICENodes("sphere.polymsh.ICETree.initPMVC.cage", >>>> "sphere.polymsh.ICETree.SceneReferenceNode.outname") >>>> xsi.SelectObj("sphere", "", True) >>>> xsi.ApplyOp("ICETree", "sphere", "siNode", "", "", 0) >>>> xsi.MoveOperatorAfter("sphere.polymsh", "sphere.polymsh.ICETree1", >>>> "sphere.polymsh.shapemarker") >>>> xsi.AddICECompoundNode(compoundPath+"\\PMVC.xsicompound", >>>> "sphere.polymsh.ICETree1") >>>> xsi.ConnectICENodes("sphere.polymsh.ICETree1.port1", >>>> "sphere.polymsh.ICETree1.PMVC.Execute") >>>> xsi.SelectObj("cube", "", "") >>>> xsi.AddICENode("GetDataNode", "sphere.polymsh.ICETree1") >>>> xsi.SetValue("sphere.polymsh.ICETree1.SceneReferenceNode.reference", >>>> "cube", "") >>>> xsi.ConnectICENodes("sphere.polymsh.ICETree1.PMVC.cage", >>>> "sphere.polymsh.ICETree1.SceneReferenceNode.outname") >>>> >>>> >>>> >>>>> On 17 December 2013 15:37, Alok Gandhi <[email protected]> wrote: >>>>> Will check that out, thanks Ahmidou! >>>>> >>>>> >>>>>> On Tue, Dec 17, 2013 at 6:28 PM, Ahmidou Lyazidi <[email protected]> >>>>>> wrote: >>>>>> Here is a new cage defomer based on this paper: >>>>>> http://www.wisdom.weizmann.ac.il/~ylipman/pmvc/pmvc.htm >>>>>> >>>>>> The result is very close the Harmonic coordinates deformation done by >>>>>> Pixar, but faster to compute, it's also better deformation than the >>>>>> Softimage factory one. >>>>>> To use it, just connect the initPMVC compound in the modeling region, >>>>>> and the PMVC one in the animation region, both on the deformed object. >>>>>> The cage must be triangulated! .. but once your happy with the >>>>>> deformation the initPMVC compound may be frozen , and then cage can be >>>>>> quadrangulated. >>>>>> The only limitation is that all the point of the deformee must be inside >>>>>> the cage. >>>>>> >>>>>> http://www.si-community.com/community/viewtopic.php?f=19&t=4683 >>>>>> >>>>>> Cheers >>>>>> ----------------------------------------------- >>>>>> Ahmidou Lyazidi >>>>>> Director | TD | CG artist >>>>>> http://vimeo.com/ahmidou/videos >>>>>> http://www.cappuccino-films.com >>>>> >>>>> >>>>> >>>>> -- >>>>> > > > > --
