Renderosity Forums / Poser - OFFICIAL
Welcome to the Poser - OFFICIAL Forum
Forum Coordinators: RedPhantom
Poser - OFFICIAL F.A.Q (Last Updated: 2024 Nov 24 1:33 pm)
Subject: applying morph targets wigs out the scale of the .obj
No, I haven't applied a scale yet but I would certainly want to be able to. The scale I'm at is much larger than I would want in Poser, but I'm leaving the import (Poser) and export (the software I'm using to make the models) options at default 100% scale as I work this issue out.
The weird (or just stupid) thing is, I got it to work once but I can't remember how I did it. In Poser I'm putting in a head and then expression morphs (all objs) that I generate elsewhere. I've got a Poser file that has a head set up the way I need it, but I can't figure out how I did that. That head scales up and down just fine and the morphs work just fine too. Any props I've saved so far go back into scaling kookery when you try to use 'em.
Could be a question of offset instead of scale. If you took the whole
figure into the modeler and then morphed it, the MT (at setting of 1.0)
will then tend toward to the position of the head, way up on top of the model.
This usually works: export the head only from Poser, with the "no world
coordinates, As morph target" box checked. Process it in the modeler
without ever moving the entire head. Export the result and bring it in as a morph.
If that's exactly what you did, and if you're truly sure that the modeler
is not changing scale on import or export, then I'm completely puzzled!
My python page
My ShareCG freebies
Oh, if you're not starting with a Poser head, then it's different.
What I do in that situation:
- I model the original item, then export it.
- With the original still sitting in the modelling app, I fiddle with vertices,
then export the changed version. - I bring the original item into Poser, with all boxes unchecked.
- I then hit the 'Add morph' button on the Poser parameter dialog and pick
the changed version.
This works even if there is a scale applied to the export, because the
original and changed versions are both on the same scale.
However, if you've turned the original into part of a Poser figure between
steps 1 and 2, then this probably won't work.
My python page
My ShareCG freebies
Okay, thanks everyone for all the suggestions
1)Yes, I've compared them in several programs, including Poser and Lightwave, they are the same object (except for fiddled vertices) same scale, load at exactly the same position, everything.
- ditto
and that's why I'm at a loss, it might as well be voodoo I can't make any logical sense out of it.
The only thing I'm sure of is that it's repeatable...it happens every time. It's just something dumb I'm doing probably, but...
It's probably not something dumb you're doing... It's probably something dumb
your modelling app is doing, like rescaling without telling you.
And THANKS!!!!!!!!
My python page
My ShareCG freebies
No, hey, THANK YOU!
Case in point, I just downloaded almost everyone one of your free Python scripts...holy cow there sure are a lot of them.
Yeah, you know this thing sure is a puzzle! I just used the same modeling app to export to LWO and LWS (lightwave object and scene), and it shows up there, with all the endomorphs auto loaded into the morph mixer, working just fine and dandy. Maybe I'll reinstall the app...
Oh, and it can't be scaling because remember, the objects are the same size when you import them side by side. It's almost as if it's the strength ( parameter dials) that's doing it somehow, because the stronger you make the effect the more the head grows large (or inverted and large for negative direction).
Quote - Oh, and it can't be scaling because remember, the objects are the same size when you import them side by side. It's almost as if it's the strength ( parameter dials) that's doing it somehow, because the stronger you make the effect the more the head grows large (or inverted and large for negative direction).
Just a couple of coppers.. I've noticed that "Morphforms" can give one headaches. Are there any V4/etc "Morphform" body morphs enabled on either the target object or when you exported the original prior to editing it?
Why don't you try this just for sh*ts and giggles:
-Export from Poser
-Import to modeling app, do nothing...then export
-Import back into Poser and apply morph
What happens?
If the morph applys with no ill effects then you know that your modeling app is not causing you problems, if it does weird out then you know your modeling app is changing the obj somehow.
Make sense?
Comitted to excellence through art.
Darkedge, thanks, yes that's the kind of testing I'd usually do, good common sense stuff...
but it doesn't apply in this case because the modeling app is not a general purpose modeling app. It only generates heads, and it starts with it's own mesh, and you cannot import an object.
I'm getting help from the app tech support, but we haven't made any progress yet, it really should not be happening. Next step: reinstall.
Quote - Darkedge, thanks, yes that's the kind of testing I'd usually do, good common sense stuff...
but it doesn't apply in this case because the modeling app is not a general purpose modeling app. It only generates heads, and it starts with it's own mesh, and you cannot import an object.
I'm getting help from the app tech support, but we haven't made any progress yet, it really should not be happening. Next step: reinstall.
Sounds then like you aren't even creating a morph at all.. you are trying to use a completely different mesh as a morph target, which simply will not work at all.
The definition of a morph is a change of point locations in an object. You aren;t changing the location of the verts in the original object atg all.. you are trying to subnstiture a completely different mesh instead.
Way too many people take way too many things way too seriously.
Sorry, (and thanks very much for ALL the suggestions) but I'm going to have to stop responding now, PLEASE read the whole thread to see that I do understand what a morph is, and what the requirements for success are, and that I have already stated that the vertices are "kosher". I will update this thread when I have the answer. Or, I may just say *&^%$% it, and migrate everything to Lightwave, where it's working. Of course, to do that I will need to get MDDPoser pack up to speed, but that's a topic for a new thread, isn't it?
:)
re: *Sounds then like you aren't even creating a morph at all.. you are trying to use a completely different mesh as a morph target, which simply will not work at all.
The definition of a morph is a change of point locations in an object. You aren;t changing the location of the verts in the original object atg all.. you are trying to subnstiture a completely different mesh instead.
I'm sorry if I wasn't clear, or clear enough. The app automatically creates both the neutral expression head, and the morph targets, generated from the same mesh and having identical vertice order. I also tried, for shits and giggles, "attempt vertice order correction" in Poser, to no avail. Plus, the Lightwave object and scene exported in the same way from the same app works perfectly. Plus, I got it to work awhile back in Poser, I just don't know how that happened.
I think we are all out of ideas as to what's wrong, and so it's natural to want to believe that I have no clue as to what I'm doing, I understand that :)
app = Facegen.
I'm pretty sure it's not that app, but I'm going to run a few more tests to narrow down and see which app I'm going to reinstall, because this is all starting to smell like an app file corruption.
First I'm going to bring the objects to my other machine and try it on a different Poser install. I'll update when I know , thanks so much!
Sometimes Poser does weird things with the normal, UV, face or smoothing group information in an imported .OBJ. You could try stripping that info out of the morph OBJ using UVMapper, and see how it ends up - after all, you only need the vertex info.
The pen is mightier than the sword. But if you literally want to have some impact, use a typewriter
Quote - app = Facegen. ...
Ah, I think I see the problem now.
First of all, you have to have the Customizer in order to use FG with other models. Creating the new base mesh for morphing can take around 10+ hours, depending on your machine, as it cranks out the numbers.
Customizer will require you to match, as closely as possible, the selected figure's head with the stock "meanface" object. In order to do that accurately, you really need to use a 3D model app an not Poser's face morphs. You can move the stock object around any way you wish but, as with everything else regarding morphs, you can not translate your own model that you will be converting.
What this process does is take an object the sofware "knows" and then tells it that a new object matches key points on the other one in such&such a way. Then, the software will create a new object to work with and recognize how to translate the work that it generates to a new set of vertices.
It's most likely that the problem lies in how you first either exported the head you wanted to use (V4/whatever) or a problem that arose when you were trying to match up your head with the programs base mesh "meanface" object. Unfortunately, there isn't any easy way to test different methods because it takes so long for the conversion process. I haven't looked at that program in ages but, unless they have made some sort of significant changes, that's where your problem lies.
NOTE - I remember reading somewhere that people had scaling problems no matter what they did. I don't mean "exploded model" problems. I mean scale problems with they exported their new morphed mesh. ie- gigantic meshes You may want to take a look for that in a keyword search. There IS a thread here or on DAZ regarding Facegen that has good tips. I never felt like investing the cash in something that didn't have some sort of better support for Poser-specific applications since that is the only use I would have for it. I read too many questions on various boards with no good answers/tutes/discussion about solutions. Most likely, most problems involve the Customizer and how it is being used (Operator Error) converting a new base mesh (V4/etc) into something FG can use to make morphs.
You can try the newest version of FaceShop Pro.. but, from everything I have seen, it produces less than stellar results unless you spend an inordinate amount of time perfecting your FaceShop leet skills. The best method is just going to be using a 3d modeling package and doing it by hand until either Facegen incorporates a Poser-Friendly inclusion (It would be extremely easy for them to do) or FaceShop begins to get more accurate results with not as many weird errors.
If you have gone through the Customizer process and are still getting these weird results, then... You may have to redo it. Them's the breaks, sorry.
If you haven't gone through the Customizer process ( a separate software package, pricey in itself) then... well... there's your problem. Time to cough up some more money.
PS - I've been working on making 3d "likeness" morphs in Hex and, it's really not as difficult as I thought it would be. Granted, I've only been goofing around with that idea for a week or so but, I can see how easy it would be once you got some experience doing it. The most difficult problem is expression morphs and having them flow well with your new face morp[h. That takes some planning on how you're going to position the verts. I don't see how Customizer would be able to do that and it would have to export its own expression morphs. (Does it do this?)
svdl, UVMAPPER sounds like a great suggestion I will have to look into that !
Quote: First of all, you have to have the Customizer in order to use FG with other models***.
Sure, but I'm not using FG with other models!
I'm using FG models. I'm pretty sure, that sort of makes pretty much everything you just so interestingly and helpfully put forth... based on a moot point. Thank you, though, and apologies if I have misunderstood you. It's definitely not my intention to be dismissive of the good info, and effort you are putting out to help me.
Listen everyone, I will update you when I get to the bottom of this, I promise.
The following summary may help us really understand what's happening:
Facegen is a package made especially for exactly what I am doing. I am not using another package, including Customizer, nor should I need to. This process (FG to Poser, with morphs) has worked fine before for me, at least once, (and I can prove it ha ha! with a poser file on my machine, an FG generated head of Clint Eastwood with all the morphs, working perfectly, done by myself) and so strange as it may sound I have confidence that the fix is some little thing, I (or we) just gotta find it (again). With my memory, I should take better notes :)
The head models are made entirely and automatically by FG. The neutral expression and all the morphs are always based on the exact same mesh, the vertices are always the same order. I am only bringing into Poser, nothing out of Poser, ever. And nothing into Facegen, ever.
Thank you all very much for your input! Ain't learnin' fun?
Okay guys, here's the skinny, and please feel free to chastise me once you learn the gruesome truth. Remember when I said I did it successfully before, so I knew it was a little thing?
Okay, it turns out that I actually answered my own question on another forum (DAZ) last december, and STILL forgot what the answer was this time around. It turns out that you just need to untick EVERYTHING in the options when you import the .obj file into Poser.
Now, when you do that, the head will be HUGE, but then you can scale down (with some crudeness, see below), and the morphs work just fine :)
weirdly, changing the export scale in FG doesn't change anything...it will still come in with the same "hugeness" in Poser...I'm not happy about that, but I'll take what I can get at this point. The problem is in accuracy, once you're down to the single digits (4% is roughly the size of a poser head) you can only be 4 or 5, not 4.5.., so your dealing with.like 20% increments if you see what I mean.
So, I'll pester FG support a bit and see what's going on with the scale thing but for now I am a happy, (and apparently prematurely senile at age 49), camper.
Hey you gotta understand, I'm juggling a lot of stuff...
kind regards, kids!
=E
Quote - *...*Sure, but I'm not using FG with other models!
I'm using FG models. I'm pretty sure, that sort of makes pretty much everything you just so interestingly and helpfully put forth... based on a moot point. Thank you, though, and apologies if I have misunderstood you. It's definitely not my intention to be dismissive of the good info, and effort you are putting out to help me.
Could have gotten to that bit a lot faster if you had stated exactly what you were using and how you were using it in the first place. That kind of info sort of helps. ;)
Quote - ...
Now, when you do that, the head will be HUGE, but then you can scale down (with some crudeness, see below), and the morphs work just fine :)
weirdly, changing the export scale in FG doesn't change anything...it will still come in with the same "hugeness" in Poser...I'm not happy about that, but I'll take what I can get at this point. The problem is in accuracy, once you're down to the single digits (4% is roughly the size of a poser head) you can only be 4 or 5, not 4.5.., so your dealing with.like 20% increments if you see what I mean.
So, I'll pester FG support a bit and see what's going on with the scale thing but for now I am a happy, (and apparently prematurely senile at age 49), camper. Hey you gotta understand, I'm juggling a lot of stuff... kind regards, kids!=E
Use a third-party scaler like Objaction Scaler (free.) Scaling accurately inside of Poser is.. problematic.. and a real PITA if you have to do it repeatedly to exact specs. Using objaction scaler you can get exactly what you want and set up a good workflow instead of spinning dials in Poser. Poser objects are scaled weird anyway. They're infintessimally small, get "soft-scaled" up inside the software because even it can't see the stuff without binoculars and Poser's Units are a creature unto themselves. I've had too many problems trying to import&scale items into Poser to ever trust it. Object centers going whacky, forced to tweak dial limites, yada, yada, yada.. Use a third party scaler and you'll be happier.
PS - Grats on finding a solution btw!
You could use an OBJ scaling utility, like Objaction Scaler (freestuff, by Maz) or ObjectScaler (freestuff, by me) to scale down the exported Facegen .OBJ files without losing any precision.
Just make sure you use the scaling utility on the exported morphs as well, with the same settings.
The pen is mightier than the sword. But if you literally want to have some impact, use a typewriter
Quote - Eyeorderchaos,
Not trying to discourage you but Zbrush can do what you want to do. It's a great proggie for morphs, textures, displacements, normals, etc. And yes, it can paste a face on a object.
Just a thought.
I'm gonna get me that program one of these days.. /sigh
But, I'm determined not to get it until I have mastered creating morphs/sculpting by hand in a regular 3D application before I feel I'll be ready to move up to ZBrush. ZBrush can do everything I want to do in regards to that, 'tis true. But, I want to be worthy of it first. :)
Thanks for the sentiment but I'm not discouraged at all, everything is fine now and I'm doing everything I want and need to, and it's working great and looking groovy.
I've played around with Zbrush a fair amount and I fairly understand what can be done with it but I won't be going that route anytime soon. It's fun to think about, (all those cool zspheres and displacement modeling tricks, etc) but I know it ain't gonna happen, ha ha!
Privacy Notice
This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.
okay kids, I've got morph targets that I know for a fact are the same scale (as the original) and have kosher vertices, and they seem to load fine but when I apply them (enter a number or move the dial) they make the obj change scale. I'm tired of struggling with this...any clues as to WTHIHH?
Thanks!