I may have figured a way to move morph targets between figures with diffrent vertex counts. I havn't tried this yet but it seems that it may actually work. What inspired this is that there are some extra vertices in the Poser Head that really tick me off. I'll use the head as an example. The two figures must line up exactly. Save the new head as an obj. Save the default head as OBJ. Set the Morph dial for one morph to 1 and export as OBJ. Repeat for each desired morph. Use MAZ's OBJ mover. Create a diffrence file for the new head from the Original. Load the PCF diffrence file and instead of using the original seed, use the Saved morphed OBJs to create the new OBJ. Repaet for each morph. Attach the new head however you normally would to a figure. Load the OBJ's created with the Mover as morph targs. Have I had too much eggnog, or do you think this will work.

Oh, now I don't understand ;-) What problem are you trying to solve using this Idea? By saying "Vertex Counts" do you mean "Vertex order"?

I understand. You are using mover to bridge the gap between P3 Figures and P4 figures....... I am going to try this later. It may just work. -Trav

Wasn't something similar posted at the PFO a couple of months ago? I can't recall clearly.... Has anyone tried the above yet? I've been wondering about it all day.

Nope, it won't work. The OBJ Mover won't convert the file back because of a mis-match. I still think I'm onto somthing here and as soon as time permits I'm going to study it more carefully.

Nerd- I tried an experment awhile ago, I wanted to delete the eyebrows from the P4 woman's face (I can't stand them, I'd rather paint them on the map), after pulling the polygons for awhile I said this is crazy (they are really close togeather), and I just cut out the whole area and bridged it with new mesh. I new Poser wouldn't accept it as a morph unless it had the same vertex count, so I cut polygons till the count was the same, and guess what, Poser took it! But as soon as I moved the dial.... As you may have guessed, her head sort of exploded. I now under the impression that the morph vertexes are referenced by their location in the object file, which, now that I think about it, is the only way it could work. But it was an intresting experment, all the same.

Yeah, it seems that the order is just as vital as count. But... But there has GOT to be a way to do this. If I can figure it out exactly, order, count. There should be a way to pull it off. Here's my thinking: A utility that compares the meshes. Old object, New object, and MT. First it will figure the diffrenct between the new and old mesh. Then take that diffrence and apply it directly to the MT. The result should result in a MT to work on the new mesh. This theory is what led me to think the OBJ mover could do it. In fact if my theory is right the Mover may be tweaked to do the job. HELLO MAZ! are you listening? or anybody with a comprehensive understanding of the OBJ file format and the way MTs work. nerd

That is exactly what I'm working on, but the task is really enormous. It involves interpolation of one mesh with another. If the verticies don't match up, find the three closest verticies, interpolate the location of the new one, and you've got it. Simple in theory, a pain in the rear in practice. Ciao! Joe

Hi nerd, Listening loud and clear. Unfortunately Mover doesn't understand meshes or vertices, it just compares two files on a byte by byte basis. When I've recovered from Christmas I'll see if I can give some thought to the problem, but don't hold your breath. It would have to be a completely new program, and may well turn out to be impossible. Even if possible it may be too difficult for me as I don't know the mechanism that Poser actually uses when morphing. It's great to see all the creative thought that goes on around here though. Cheers, Maz

You guys are definitely on the right track, it would save me ahell of a lot of conform, morph work in max to be able to transfer p4 targets to my new model this way, I think that the key may be to narrow the comparison list, ie select the orignal target area vertices only. and make a reference list of all vertices in the selection as they pertain to the .obj File, then Reorder the verticelist of the target object with the closets matches, you would certianly get some errors, but even if you got 90% clean up wouldn't be too hard, and it caould all be done in ascII. Maybe our friend Anthony might be willing to tackle it since he burns that stuff out so damn fast (bless his heart.) I wouldn't mind writing a Maxscript like that myself, but I'm way behind right now look for My coming post thought! peace,