Oh my gosh, no, I did not. :x I was following a checklist of stuff to do after making the lods with DirectX and it didn't mention it would need to be renumbered. I've done that now and it looks like it's working correctly now - no more error messages at least, and the BGEO appears to be generating properly. Should have occurred to me, really, since that DirectX tool seems to wipe out everything.

In any case, this was the error message I was getting before:



Not sure if you still want to see the problem files because renumbering has fixed the issue, but I attached a zip with the simgeoms of the old version of lod2 and its thin morph.

ETA:

Is MorphMaker currently the only program that can add morphs to packages? It absolutely refuses to work for me; every time I click the "add morphs to package" option, it just freezes for a second and then nothing happens.

I need to put a check in for this kind of problem with an error message that actually tells the user what's wrong instead of confusing them with array bounds nonsense, and your files should be useful test data. Thanks!

Currently, yes, MorphMaker is the only one. I think you mentioned you're using Windows 7 - have you updated to the latest version of MorphMaker? The latest version of the helper library (salflib.dll) seems to have made some big improvements to the functionality under Win 7.

Making an entirely new version using the same interface as Toolkit is now #2 on my list so I should be working on it in another day or two and how long it takes depends on how long it takes me to figure out how to use the S3OC libraries - shouldn't be much of a problem.

No problem, glad they'll be helpful to someone! I'm glad you'll be adding the function to the Toolkit, it's nice to have it all in one convenient program.

Yeah, I was using the most recent version of MorphMaker. I've redownloaded the program though, in case there was something wrong with my initial download, and I have it set to run on XP Compatibility as well and it seems to be working alright now. Must have been something weird going on, but hey, I'm not one to look a gift horse in the mouth.

However, still issues with the morphs ingame...



There's a neck gap with all sizes, but it starts deforming like crazy once I get close to maxing out the weight slider. The hair seems to deform with the body too. All of the morphs looked okay in Milkshape, nothing weird at all.

Wow, weird! Could you also upload the package that's giving you this effect?

Did you match the neck seam with the head in Milkshape? If so, it could be a bone assignment problem - you could try the Toolkit seam bone assignment matching tool in the Bone Tools tab and use the head as reference. What number did you use as a start number for the renumbering? I'm wondering if you're overlapping with the hair range? This isn't a super high poly mesh, is it?

Sure thing, attached it.

When I initially did the meshing for this, I took the neck right off one of the extracted reference meshes and attached it, so it definitely matched up exactly with the head. The current mesh in there right now is the fixed mesh Bloomsbase uploaded for lod1. I noticed the neck seam earlier when I checked that mesh out ingame, but we both thought it would be fixed once I added morphs. The neck looks to be weighted correctly (50% head, 50% neck?), and I used 5000 as the number to start at for renumbering, because it was the one listed for tops under the Hint button. And no, it's not a high poly mesh. I think it's pretty low poly by ts3 standards, actually, it's 1623 polys at lod1.

ETA: Just did a quick check of Bloom's mesh in Milkshape in case something weird happened, but nope, the neck seam still matches up perfectly with the head reference mesh.

i removed the neck completly and replaced it with the one of the aftopnaked, should be ok
Must be something with the numbering, did you do it in Milkshape or does have toolkit a option for it aswell?

In lod1 you had four vertices with vertex IDs of zero. I fixed that and replaced the lod1 base mesh and the BGEOs, and now the hair doesn't distort.

IMO you need to recheck your morphs - when I import the lod1 fat and thin morphs in Milkshape I get the same neck distortion as in the game. Bloom probably has the right idea, you can Frankenstein the top with the neck from an EA mesh. That would probably be easier than trying to fix those morphs.

Also you might want to check the bone assignments on the breasts - below is a pic with the breast size slider maximized. And you have some clipping issues around the hips in the morphs.

Bloom, I renumbered with the Toolkit, there is an option for it.

Thank you for the fixed files! How would that have happened? o: The vertex IDs being zero?

And that's really weird, because I wasn't seeing any kind of neck distortion when I checked them earlier. Could it be something that's happening on export, or when it's turned into a BGEO? Because I'm looking at the morph GEOMs again in Milkshape and they look totally fine. There's no neck weirdness or mismatches. But maybe I'm not understanding the problem correctly.



So if I do need to start over with the morphs I should add a neck from an EA mesh to the base mesh, and recreate the morphs? Or take the morph GEOMs I already have and add an EA neck to them?

Yeah, the maximized breast slider is something I noticed sort of belatedly, I believe it does that because there aren't enough verts in the chest area to morph into a proper breast shape (due to the mesh being a lower poly ts2 one, which was originally a male mesh that someone converted to female). I was debating whether to add some more edge loops or something there, so it deforms into a more rounded shape better. But I'm no expert at modeling yet, and my prof always tells me that you have to keep the poly flow of the mesh consistent (something adding edge loops would disrupt) so I wasn't sure how to go about fixing it properly.

I generated morph geoms from the BGEOs in your package, and when I imported them into Milkshape over the base mesh they had that distortion at the neck. Stupid question time: Are you sure you're looking at your latest morph geoms imported over your latest base mesh and that the latest morph geoms are what you used to make the BGEOs? The picture you uploaded seems to be the base mesh although it's hard to tell, and the base mesh isn't the problem. Did you possibly renumber the base mesh or the morphs separately at some point? You have to load the base and all the morphs into the renumber tool at the same time for the numbering to be done exactly the same for the base and all the morphs. If the numbers aren't the same, the morphs won't work. If that's not it, I'm kind of mystified what could be going on here - could you upload your set of base and morph geoms?

Oh, and the zero ID vertices - the usual scenario would be that you added vertices in Milkshape and then forgot to renumber. Can't say for sure, though.

Edit: Adding faces to the breasts would involve more work - fastest way would probably be to split faces in the base mesh, use the Toolkit auto-bone to assign bones to the new vertices, then use auto-morph to make new morphs. Actually since you need to rework the morphs anyway it might be worth doing.

Never a stupid question, sometimes the most obvious things are the problem! But yes, I'm positive I created the BGEOs from the latest morph geoms, I have no other geoms I could have mistaken for them. And no, that's the thin morph in the picture I have uploaded, and the neck is fine for the fit and preg morphs as well. The only one that doesn't match up is the fat morph, and that's to be expected (I think?), since I'm using a reference mesh for the default weight AF. I renumbered the base mesh itself before I created the morphs, but since then I haven't renumbered it at all.

That would be odd, I hadn't done much of anything to the mesh in Milkshape besides use it to convert my original .fbx into a simgeom, so I dunno where four new verts would have come from.

ETA: That's a good idea, I'll try that. Do I just use the subdivide option in the Faces menu of Milkshape? I don't particularly want to import back into Maya just to split a few faces. This doesn't mean I need to rework the lods too, does it?

Possibly the zero vertices in the base and all the morphs are the problem - all I did was renumber the base and morphs and remake the BGEOs, and now the neck looks perfect in the morphs. Having those zero ID verts created a big list of non-data vertex IDs in the BGEOs, and it's possible the game went outside an array or who knows what. I really have to build in a numbering validation and a warning about that.

BTW, if this is how you usually name your meshes, my advice is to name them in a more identifiable way with the lod and the specific morph and maybe a version. With 'morph 1', 'morph 2', etc. with dozens of meshes flying around you're going to mix them up at some point.

Yes, Faces / Subdivide works, although personally I prefer Vertex / Divide Edge because I have more control over how the faces are divided and it doesn't give me vertices floating on an undivided edge which results in gaps in the mesh when morphed. (Select two vertices and divide the edge between them, with the faces on both sides of the edge being split.) With Divide Edge and some of the Subdivide Faces you'll get new vertices with bone weights already assigned using a weighted average so you may not have to fiddle with them much. And no, no need to redo the lower lods - they're supposed to be lower detail anyway.

Oh that's great! It does look awesome ingame, and it indeed morphs properly now. But the breasts are still an issue, so I'll have to reconstruct the morphs either way.



Is it just me, or is there a small texture mismatch at the neck? The other side of the neck is seamless, but this one appears to have a little tiny line. Am I just nitpicking/seeing things at this point?

Haha, I am pretty disorganized with my file naming usually, baaad habit, but I know those ones are lod1 because I named all my other morphs lod2_fat etc.

I'll try out the vertex/divide edge tool for sure then, thanks.

ETA:



Back to the drawing board...

the breast on sim 2 meshes are very blocky on some meshes, including this one.
Dividing the faces and reassigne the new vertex will smooth them out

Yes, will do that, thank you. School is picking up a little for me, so this project is going to have to go on the backburner until I have some guaranteed free time to work on it more. Thank you so much to the both of you for your help with this, I've learned a ton. It was really helpful to have such great guidance. :lovestruc

Oh man, I really hope that someone is still watching this thread. :x

I was just going to give the person I was creating this mesh for the fixed mesh with morphs, and continue with the breast slider thing once I'm on school holidays again, but I noticed some really terrible clipping issues with bottoms that aren't on nude meshes (stupid me for not even remembering that the pants I was testing with were skintight). Anyway, I fixed the clipping and decided to use TSRW to rebuild the package this time because CTU was being uncooperative and throwing errors, and ingame now the hoodie looks fine, but all of the bottoms' textures are layering over it, and they weren't before. I thought it was because I cloned a tank top instead of the long tunic I was previously using as a base, but going back to the tunic just gave me ugly normal issues and still everything layered over it. Is this a TSRW thing? And is there a way to change the layering so that the top texture layers over bottoms as it should? I just want this mesh to work fine and go away.

Edit: Running a new package through CTU when basing it on either the tank top or the tunic now also gives me the same layering problem. I didn't touch the uvs when I fixed the clipping, and a cursory glance shows no change in the map from anything else either.

Edit 2: This is a longer mesh and thus, it does break into the territory on the uv map reserved for bottoms, but there was previously no issue whatsoever with any overlapping textures. I based the placement of the uvs off some of EA's longer meshes which also layer over bottoms and edge into the bottom map area a little bit.

It depends wat you pick first in CAS if i remember correct

Just tried that, and sadly, it didn't work. Whether I pick the pants first and top second or vice versa, it still overlaps. I'm really confused about what could have suddenly changed to make it do this.

as far as i know it always has been like that, sometimes it works in CAS by accident till you change outfits again
You could try playing with the part mask as that is the texture that ''should'' control this(but with CTU i never got it to work correct.)

I never had it do it with all of the other versions of the hoodie; and I did the exact same thing every time I stuck it ingame for testing. Picked the top, then the bottoms.

I'm using TSRW for the moment, so maybe I'll have more luck with messing around with the part mask than I would using CTU.