as though I'm making too many concurrent [Request] and "help" threads as of late, but I've got, like, a month left, and these things have just been sitting around. Shame on me for procrastinating. :/ Still, I'd like to think I've done a good amount myself so far. Anyway...
This is another one of those psuedo-Requests that could feasibly help other people and be a useful 'quasi-Tutorial' (similar to the "Making a Worker" thread).
I would like to create a mounted unit out of 2 separate unit models ('mount' and 'rider'). However, as opposed to just plain attaching or geomerging (both of which freeze or remove animations), I know of a method by which two actual models can be "merged", yet maintain there animations.
Method?
Quote:
Originally Posted by DonDustin
there are basicly 2 tools that can do this:
1. oinkerwnikle´s RedistributeKeys
2. the russian tool mdlvis which is also avaivable in english
the oinkerwinkle program is the 1 that works better,
but mdlvis is much faster and easier but you get some bugs ( the anims of the puppeteer do not fit 100%)
th problem with the oinkerwinkle tool is that you have to make it for every single bone for every single animation. And if you have 20 bones and 10 anims you will have to make 200 edits in the mdl.
Quote:
Originally Posted by Kyrbi0
Hmm... So which one did you use?
And, perhaps more importantly, would you be willing to do it again? I've been waiting for this opportunity for a while; no "pro-modeller/animators" ever seem available, except jigrael; and he's already done enough for me. :P.
Quote:
Originally Posted by DonDustin
It´s like that:
I used the oinker tool for my basilisk first ( should use naga seawitch + hydra anims first) but the problem was that the hydra´s movement was so strange. It took around 2-3 hours for around 7 bones.(maybe because I worked a bit slow^^)
I wanted to do the same with the puppetmaster but it was too much work.That´s why I used the easier way.
I would make it for you.Depending on how many bones the model will have I will use the oinkertool.
Quote:
Originally Posted by DonDustin
BTW: I will make that mdlvis tutorial just for you because I don´t want that some people know about that tool because it can export wow models-.-
Oh and do you have a copy of your english mdlvis in your warcraft folder?? Else you will have to work with the russian 1^^
And here is the first part of the tutorial( I will use the spiritwolf anims for the mount and archer anims for the rider):
1. you create your 2 meshes but as seperate models: rider and mount
2. you take a look at the anims to decide if some anims would look strange when both models are merged: if this is the case----> try to replace that animation with animation transfer
3. you add to both models their skeleton and make sure both look good in Magos
4. you remove unused bones with magos( make sure that you still have a copy of your edited model)
5. you make a grid like this:
sequence______rider__________________mount
attack________intervall of rider_________intervall of mount
..............
6. you decide if you put the mount intervalls as the rider anims intervalls or vice versa depending on the number of bones the models have(the 1 that have more bones will be the base model)
7. you make a copy of the model which intervalls you will edit ( name it somethink like #####newanims.mdl)
the rest will be explained in future^^
I am not claiming that 'Dust' (DonDustin over at THW) here is the sole proprietor of this knowledge, but he is one of the few active people I know of that has recently done this, and inspired me to realize it's possibilities (via his Puppetmaster resource).
~~~
So anyway, here are the models that need to be "mounted".
~Warbringer: Skullsplinter + Dinosaur~ +
Permission to Edit
(Norinrad, Skullsplinter)
Sure no problem, yo have my permission ^^. Sorry for late response but I'm not checking hiveworkshop very often.
~Marauder: AxeTroll + Raptor~ +
Permission to Edit
(levigeorge1716, AxeTroll)
Sorry for the very late reply, but ya, go ahead.
Part of me thinks it would be much easier to paste on an unanimated rider (including its skeleton) and then animate those parts from scratch in Warforger. Of course key redistribution is always an option, but as Dust said, it's slow and you'll have wasted your time when the motions just don't work in a mounted context.
Part of me thinks it would be much easier to paste on an unanimated rider (including its skeleton) and then animate those parts from scratch in Warforger. Of course key redistribution is always an option, but as Dust said, it's slow and you'll have wasted your time when the motions just don't work in a mounted context.
Hm. So you think I should just geomerge the two things and post the .mdl for that here, and we can go on from there? (then again, I don't know how I could include the skeleton that way).
I think I have Warforger... But would MDLVis be any better?
I know nothing about MDLVIS, but I have worked out how to get Warforger operating correctly without noticeably demolishing existing animations, so I'll stick with that.
Oink's Geomerger is actually a false economy. You can achieve the same result simply by copying and pasting the geoset in notepad, plus it preserves the vertex groups. Then you'd have to slot in a replacement skeleton, hook up those values into the groups over the old ones, shunt it all around so the man's on top instead of the bottom...
Alright. Well, although the 3rd option is actually the least important to me, I guess I'll start with that since that's the most finished (#1 needs some TC on the lizard and plausibly some armor, #2 is most important but the raptor has no bones, anims or TC).
So I take the two models (heretofore referred to as "Troll" and "Spider"), make them .mdl's, open the Troll in Notepad, and copy/paste the whole thing (or just the body?) into the Spider's mdl file (also open in Notepad), to start.
--EDIT--
Eh... I copy/pasted all the geosets of the Troll into the end of the Spider's mdl file, but it would only let me save as a Notepad document. I've missed something integral, haven't I?...
Well, so it's been quite some time. Is there... a "next step" in this process? Or perhaps a tutorial and that dummy model so I can attempt it myself? Or... what?
