Soulcage: Six magic letters: OBJ - MDD

[Elm]

One more thing:

 

Have an eye on the model's normals - they should all be aligned outwards (or 'correct') - as modo does not interpret a face double sided by default (you can of course tell modo to do that, but correct normals are always better...)

[3DArtZ]

Even the girl in the Lefax ad? Just pose sliders for the face?

 

You really don't need facerigs... they are often more limited then good poses which can be blended and added to each other much better.

I only use small "rigs" for eyes, maybe for the nose (for dynamic constraints, if it is useful and necessary) and the chin...

Everything else can be handle with poses in a less complicated and nice way. But that is just a matter of taste...

 

See you

*Fuchur*

I've been waiting a long time to hear someone else say this...

[Vertexspline]

Well....I have tried a test simple blob into am from modo- exported obj from Modo--imported obj into am --placed a few bones ...moved them a little and exported the obj and mdd from am --imported to modo the obj and then mdd and yep--- it works!!

 

Now I have no experience with mdd and modo before so trying to test some more things but so far ----pretty interesting indeed. thanks elmar.

[kikiriki]

Thank you soooo much for this! :-) I wasn't aware that A:M supports MDD, This opens so many possibilities. I haven't been so excited for years... really!

 

Btw. Elm, you mentioned that you could do Hair in A:M and render it elsewhere. Can you please explain how?

 

Oh... I forgot to say: Excelent work!!!

[Elm]

...

Btw. Elm, you mentioned that you could do Hair in A:M and render it elsewhere. Can you please explain how?

...

 

No, I didn't mention that. I wouldn't have any idea of how to get this done. That Santa image Mike posted wears A:M hair, whilst the animated commercials Santa's got XSI hair. These have nothing to do with each other, hair-system-wise.

 

...

I wasn't aware that A:M supports MDD, This opens so many possibilities.

...

 

Thanks to (mainly I guess) Steffen Gross!

[jakerupert]

>QUOTE(Fuchur @ Nov 17 2011, 09:30 AM)

QUOTE(Gerry @ Nov 17 2011, 07:33 AM)

Even the girl in the Lefax ad? Just pose sliders for the face?

 

You really don't need facerigs... they are often more limited then good poses which can be blended and added to each other much better.

I only use small "rigs" for eyes, maybe for the nose (for dynamic constraints, if it is useful and necessary) and the chin...

Everything else can be handle with poses in a less complicated and nice way. But that is just a matter of taste...

 

See you

*Fuchur*

 

I've been waiting a long time to hear someone else say this...

 

 

 

but that might change in the future...

[jakerupert]

So I asume the MDD workflow would work with any other 3D app also like C4D or Blender for instance?

[Elm]

So I asume the MDD workflow would work with any other 3D app also like C4D or Blender for instance?

 

 

It does work the same way in XSI at least, so I guess it also does in any package that supports mdd files.

[mike meyer]

Oh... I forgot to say: Excelent work!!!

 

 

Thank's a lot for the kind words! Good to see you're still around, using A:M.

 

 

 

I'll try to post a few more pics of the santa the day after tomorrow.

[jimd]

just caught this

wow! that's so good, wonderful work

j

[jakerupert]

As far as I understand it, the only reason for this workflow is the good and fast rendercapability of Modo, right?

 

So would it be possible for Steffen to lift A:M`s build in renderer to a similar level or would that be too big a task for

one single man?

[Elm]

As far as I understand it, the only reason for this workflow is the good and fast rendercapability of Modo, right?

 

So would it be possible for Steffen to lift A:M`s build in renderer to a similar level or would that be too big a task for

one single man?

 

 

Primarily, rendering, yes.

But there's more. Like Special effects for example - cloth, fluids, other things... which just ARE more advanced in other packages.

[ernesttx]

Ok, finally got .mdd working for me. The weird thing was trying to figure out that you have to make sure your FPS are set. Another weird thing I'll have to check out is that I couldn't directly import into Vue 10. I first had to import into messiah and then export out the .mdd from there for Vue to read it. Also, I think that the Python camera export might need to be look at as I could not get a camera imported into Vue with the keyframes. I looked at the Python code and an example from Vue and had to change the first 3 lines before the camera could be imported.

 

Other than that, everything went smoothly from A:M to other software. Blender could read the .mdd file fine. Anyway, made a quick test animation, enjoy:

 

http://amfilms.hash.com/video/1995/Thom-Run

[robcat2075]

Also, I think that the Python camera export might need to be look at as I could not get a camera imported into Vue with the keyframes. I looked at the Python code and an example from Vue and had to change the first 3 lines before the camera could be imported.

 

Could you post the specifics of what you had to change in a thread here? i know another user who was stuck on that.

[ernesttx]

Yeah, Rob, I was planning on doing a tutorial on all of it. I spent the better part of 3 days trying to figure it all out hehe Being the TD and the artist is a painful thing.

 

Anywho, long story short, when you export .py from AM these are the top lines in the python code:

 

DeselectAll()

cam=StoreCamera()

SelectByType(VuePython.VUEBasicObject.VOT_Camera)

cam = GetSelectedObjectByIndex(0)

 

And it wasn't importing into Vue with the keyframes. It would import and create a camera. I do not know anything about Python or coding, but I do know if something else works, it's all about trial and error and find what does work. Now, it may be that with Vue 10, somethings have changed in the API and some parameters are named different and all that. Then, I looked at the Bouncing_Sphere.py that comes with Vue, and found similiar lines and changed those. These are the 3 lines that I changed in the AM .py file:

 

DeselectAll()

SelectByType(VuePython.VOT_Camera)

cam = GetSelectedObjectByIndex(0)

 

Save new .py file and try to import. Voila. Now, I don't know alot so you may have to put in the cam=StoreCamera() line as well. All I was trying to do was get the file into Vue. So, any other help or information would be great as well.

[ruscular]

Have anyone tried it with Messiah3D animate? If not I'll get back on this weekend to get it tested with my new MDD plugin for both Messiah3D and Lightwave3D.

[ernesttx]

hey ruscular, the .mdd export from AM works with messiah. I just used the DMorph effect to load the file. If you have a new plugin I would be more than happy to test out as well. I have Messiah 5 64 bit. Win 7.

[Elm]

Great, folks!!!

[ruscular]

It work! I was able to import obj file into Messiah and the MDD file very well! Which means I should be able to add better hair. Also I prefer the rigging with smart skin for body dynamic and mouth in Hash.

[ruscular]

http://www.cyberdreamcreation.com/images/SLTestlowBK.mov

 

Here a few thing I learn today! I can reposition the model to put it anywhere I want, even after applying the MDD file to the object inside Messiah. I was also able to resave the MDD file inside Messiah and a 7 second MDD file went from 800mb to 93mb. I was not able to move the character inside LightWave3d once the MDD file was applied. The only thing I had to do is re-assign the shader material to the model surface area.

[John Bigboote]

DIGGIN that test... not sure whats-what betwixt A:M Lightwave3D and Houdini,,, but I am DIGGIN the test!

[ruscular]

DIGGIN that test... not sure whats-what betwixt A:M Lightwave3D and Houdini,,, but I am DIGGIN the test!

I thought about you! Because of the hair challenge. Messiah has hair particle and I am not sure yet on how to apply hair to a obj file once hair is applied afterward. Create a null to the point on the obj, and then apply a scalp mesh to the null?

[detbear]

>(yes it looks a bit crappy but you shouldn't care much)

 

 

Because the cp's topology from modo stays the same in A:M - which is a bit confusing to A:M's spline geometry (looking odd here and there), but once you export the mdd data from A:M and feed the modo-model with it, everything looks as it should.

 

There's one more amazing detail in this context: This mdd exchange also works the other way round - even the face poses of the lefax girl were actually modeled in modo. Then they were animated (still in modo) in a frame to frame mannor. This action was exported as a mdd file, loaded in to an A:M action. The face cp's positions were then copied from each action frame into a corresponding percentage pose of the A:M girl.

 

 

Elm....This is a late post on this thread but I have 2 important questions concerning the mdd to pose process:

 

1.Do you import the mdd into a pose and copy it there....Or where do you access the copied mesh from...and where do you paste it...

 

2.I want to import a model from other software. I can import an obj from the other App, but it of course will not bring the rig. However, I can bring in .Mdd of

the joints bent(elbow, knee, neck, shoulder, etc) If what you are saying is true, I should be able to smartskin/ pose these by some how pasting keys of the cps from the mdd file exported out of the other App into Hash. This character would of course need to be rigged in Hash prior to using the Mdds. What is the method used to get the mdd keyframes pasted into the percentages of each pose. My hope is to round about get the smartskin in the pose to morph to the morph shape I export from the other App.

 

William

[Fuchur]

As far as I understand it:

You import the MDD-data into an Action. After that you've got keyframes (in A:M) in the action. In the action you can now copy the CP-animation-data in muscle-mode, create a new pose for your character, go to the percentage the keyframes you copied from the action should be at and just paste it.

 

However you will not very likely get around rigging again, at least not for more complex motion. CP-animation is linear and will "morph" from A to B. If you need to move an arm or so you need bone-animation with joints which will create the arcs automatically... for basic morphing like it often happens in the face you can get it done with that method (very well) but for example a walking character will look very strange if you don't copy and paste each frame a keyframe.

 

See you

*Fuchur*

[detbear]

I guess, effectively...with some effort, this would be a way to import the skin weights/morph targets of a rigg from another App and use the pose process in a smartskin to morph the joints for things like elbows, knees, fingers, etc. Especially if the imported OBJ character has a bunch of cps that would be hard to weight and/ or smartskin. Not to mention the facial morph transfer posibilities.

 

William

[detbear]

Hey Fuchur.

 

Thanks for the reply. I think you are right along those lines. However, I'm looking at a specific part of the rigging process in the smartskin "Pose" Action. This may not even be possible or an option, But moving cps at each percentage of a heavy mesh can be daunting.

 

But if you could import mdd motion from the rigged app and paste those cp frames in a smartskin pose....well.....the work would be done for you....Again...if that is possible.

 

For instance....An mdd of an elbow bend could be used by copying 3 keys: 1 at 00% / 1 at 50% / and 1 at 100%. Then those keys could be pasted in the smartskin pose action at the same percentages.

[Fuchur]

Hey Fuchur.

 

Thanks for the reply. I think you are right along those lines. However, I'm looking at a specific part of the rigging process in the smartskin "Pose" Action. This may not even be possible or an option, But moving cps at each percentage of a heavy mesh can be daunting.

 

But if you could import mdd motion from the rigged app and paste those cp frames in a smartskin pose....well.....the work would be done for you....Again...if that is possible.

 

For instance....An mdd of an elbow bend could be used by copying 3 keys: 1 at 00% / 1 at 50% / and 1 at 100%. Then those keys could be pasted in the smartskin pose action at the same percentages.

 

Yes, but keep in mind you still have only linear morphing there... it will not look the same as real bone-animation. Anyway it would be possible for a elbow or something. You could get away with 3-5 keyframes there... for more komplex animation it would not work... anyway it could be a very nice addition...

 

See you

*Fuchur*

[detbear]

I did a quick experiment that resulted in the imported mdd crunching the whole Hash model.....really ugly.

 

I moved an elbow bone in a Hash model in the Choreography. Exported the data as mdd

 

Then imported the mdd data back onto a boneless version of the hash model(just to see what the results would be)

 

I must be missing some key elements in the process.

 

I found that you must import the mdd onto a model in an action or choreography. The result shows the move across

the keyframes. But their are deformations and the model is all messed up.

 

A step by step process guide is needed in this case for me at least. The import of the .mdd gives an error also.

 

I'm using v16 on a pc

[ruscular]

For the smoothest looking model I check normals and export with triangles and at 4 polygon per patch! I also imported it into Lightwave the mdd file and then resave it and reduce the mdd file. The mdd file export contain too much unneeded info. I did not resaved the obj file as that resulted a mess up mesh when imported back to Messiah renderer.

 

I exported the mdd from choreography 4 poly per patch w/triangle and the model to OBJ file 4 poly per patch from AM:

import mdd to lightwave and resave (not require but a preference)

load the obj and resaved mdd to Messiah, and fix material setting and adjust subsurface scatter setting on skin and render final output

[detbear]

sounds like a good workflow to Messiah.

 

I wonder if your process would work back into A:M instead just as a trial.

[ruscular]

sounds like a good workflow to Messiah.

 

I wonder if your process would work back into A:M instead just as a trial.

Pardon if I understand correctly, but make a obj in Lightwave and then export it out to AM and then rig it and animate it, and then export it to Messiah?

[ruscular]

Object and MDD file export from Hash also work inside Poser! I wouldn't dare import a poser model into Hash, but maybe that retopology feature will come in handy for that?

[Willi]

2.I want to import a model from other software. I can import an obj from the other App, but it of course will not bring the rig. However, I can bring in .Mdd of

the joints bent(elbow, knee, neck, shoulder, etc) If what you are saying is true, I should be able to smartskin/ pose these by some how pasting keys of the cps from the mdd file exported out of the other App into Hash. This character would of course need to be rigged in Hash prior to using the Mdds. What is the method used to get the mdd keyframes pasted into the percentages of each pose. My hope is to round about get the smartskin in the pose to morph to the morph shape I export from the other App.

 

make an action of your imported model

expand the action, so you can see the "shortcut of your model"

be sure "animated mode" is on

right-click on the "shorcut of your model"->plugins->import mdd

now you have an "muscle"-animated action.

 

go to the frame you want to copy out the keyframes. select cps and edit->"copy keframes"

make a new percentage pose on your model. select all cps (or only those that should have a pose-keframe), edit->"paste keframe"

voila. you shoud now have a percentage pose...

[detbear]

make an action of your imported model

expand the action, so you can see the "shortcut of your model"

be sure "animated mode" is on

right-click on the "shorcut of your model"->plugins->import mdd

now you have an "muscle"-animated action.

 

go to the frame you want to copy out the keyframes. select cps and edit->"copy keframes"

make a new percentage pose on your model. select all cps (or only those that should have a pose-keframe), edit->"paste keframe"

voila. you shoud now have a percentage pose...

 

 

Willi....

 

Thanks so much for responding. This is a process I've been wanting to learn how to accomplish but the proper steps have been unclear. I will try it out.

I'm hoping to "Backdoor" a rigged character from another app that way. My original hope was to use this to get the CPs moved in a kneebend into smartskin somehow.

 

But there are sooo many other uses for this process also.

 

Thank you for responding.

 

William

[chucky]

Awesome, looks great, tastes funny!