Such a BIG thank you for this script to ColorCurvature.

Why the excitement you ask (or not) Well as you likely know by now I have been working on this male figure for about 3 years now.

I am down to loading JCM's. I spent the last month perfecting some of the most difficult JCM's in the figure, for the thighs and shoulders. To my utter dismay I realized I had been loading all these morphs on the wrong OBJ. You see I had wanted the figure to load without the genitals and you could add them later with geometry insertion. I had been loading all my JCM's on the figure from the full OBJ with the gens welded in.

Hot and cold flashes to say the least. You see if I now load the figure with the correct OBJ with no gens and the hole plugged, it changes the vertex order and none of my morphs will work.

I read in the manual that ColorCurvatures PoseMorphLoader can also load a morph using the group information without the vertex info.  So I booted it up and started loading.  Worked just gorgeously. No loss of precision that I can see.

OMG, I am soo relieved. So I am just thrilled with this script. If you are ever addle brained enough to do what I did there is now an excellent solution.

cheers.

Nice way to slip in an ad for a new product, thought that wasn't allowed?

Quote - Nice way to slip in an ad for a new product, thought that wasn't allowed?

oooo cynical :)

Quote - Nice way to slip in an ad for a new product, thought that wasn't allowed?

It's not my product I dont have any thank you. Furthermore it's no different than talking about how to use some feature in Poser. I am sharing a problem experience and a damned good solution to it. Does everyone already know a better solution to loading morphs on a figure with different vertice ordering? I suppose you can tell us another method? Then why havn't you?

cheers.

PML is a must have.I was cut down to size a bit, when I realized that it cannot replace a good rigging.I did tons of JCMs for Miki2 and they work fine untill the moment I try to combine her fixed poses.I mean I fixed "arms up" and "arms front" .But when posing "arms up and front" ,I had those ugly joint creases again . If you would like to create an animation, you are forced to create lots of keyframes and fix each of them seperatly with the PML tool.To point it out clearly : the problem is the Poser rigging not PML .

Hi Timberwolf,

Ya it gets really insane when you start combining morphs for different joint rotations. That is what I am doing now with my figure. Getting the X rotation and its JCM morph to work with the Z rotation and its JCM morph on the thighs was pretty frustrating. Just for the two thighs and a couple of rotations on the shoulders it has taken me almost a month now to work out the Morphs so the work well together. The rigging of course has alot to do with being able to get the morphs to cooperate.

I leave a little slack between the falloff zones, in certain places and keep it tight in others to accomodate the JCMs I want to put in. But its still a bear..

You can understand my horror when I noticed my error with the wrong OBJ. I was about to just quit and give it up.

I have seen many a thread in this forum in the past asking how to load morphs onto a figure when the vertice order was off, In the past you could sometimes reorder the vertices with kaweckis morpher, IF and ONLY IF you had the same number of vertices in the two objects. In this case in the past it would have been a total disaster because not only was the vertice ordering off, the two OBJ's did not even have the same number of vertices.

cheers.

Hi phantom,
this is an interesting observation indeed!
This would mean that the different .OBJ you were using are sharing equally structured groups.
It might also work if one .OBJ has more groups than the other, but I am not really sure.

I think the load process would crash if it would fell over groups that have the same name but different structure, because unequal structures are not expected in the code.

Did you use the loader itself or the morph copy program that comes with it?

I'm glad the tool helped!!
col

Quote - Nice way to slip in an ad for a new product, thought that wasn't allowed?

I  for one am really glad he did it.
F*ck the rules.
I read this, and then read about PoseMorphLoader, which I had never heard of (and wouldn't even thought possible), and now I'm going to have to get it. I have been wanting to be able to do the things it claims to be able to do for a long time now, just thought there was no way possible.

So thanks for the ad, Phantom! :D

colorcurvature

I used the loader. The two OBJ's are the same except one of them does not have the genital and instead has a cod piece that is welded in. And the vertex ordering comes out diferently when the figure loads into Poser.

But I had my Morphs on body parts that were the same group and structure just different vertex order. So what I did was to export only the morphed groups as the morph target from one figure and loaded them onto the other figure using the group option instead of the vertex or auto. It worked very well. If the figures had been different figures entirely I dont think it would work.

Oh, and I know the manual says the morphloader has not been tested in Poser Pro 2010, but I have been using it in Poser pro 2010 and have had no problems.

Ya it really bailed me out this time. I have been using it alot in the last month. I had a question for you. If I save these morphs out as delta or inj files to load later do I loose the unified morph deltas at the seams or is that in the morph itself?

Schecterman:

I wasn't really trying to advertise, I was just really happy with the loader.  But Glad you found this of interest.

cheers.

I think the morph deltas can be exactly copied from the figure into an .INJ, but I must admit that I do not have experience here. So unless the used .INJ writer changes the delta values (which I do not assume it does), it should write the data exactly.

The unified deltas is a calculation method - but once its all calculated, the INJ just gets the final values.

Which tool are you using for writing .INJ?

I have been using the Binary morph editor or the INJ-REM script, they are both tools by D3D. I think from what I read the scripts just  rips the deltas directly from the cr2. So it should be exactly as written.  If the unified morph deltas are written into the morph I think it will be copied into the INJ file.

Yes. It all boils down to morph delta values.

To explain why there is unified/non-unified loading:

On body part welds, one has two morph delta values s for the same vertex (sad but true). Only one of them moves the vertex - the other one is hidden, and only has effect on the normal (and therefore the shading). Unified means that both values will be set identical - this is what most programs do, and that mostly works good.

My personal experience was that non-uniform deltas can make results with a better shading, especially if the joint is highly disrupted. But this is not always true, this is why both modes are included.

I tried for a while to use the hidden morph delta to produce normals that are perfectly smooth, but this seemed to be not possible or just outside my capabilities. The modification of a vertex morph delta modifies not only the normal of just this vertex, but also of the vertices that are around it, which makes things more difficult.

I'm really looking forward to see the JCM's you made!!