Jump to content

Muscles And Jiggle


Recommended Posts

I would wait til H8.1 comes out and then decide, if you can afford to wait.... sorry to be so criptic. Not even sure if I should be saying that. Hopefully that isn't giving too much away.

26901[/snapback]

Go on Simon you're on the right track...:):):)

Link to comment
Share on other sites

I have a couple of questions/concerns in regard with the muscle system in maya. I used maya a lot but a long time ago so I may not be up to date...

Jiggle is my main concern. In the old days alias said that jiggle was developed to simulate lag, a extension to the softbody. I believe they had an example with a fat guy walking around and the belly jiggles.

Now the problem: as far as I know, but I may be wrong, maya's jiggle is in world space, which makes it unusable with muscles whom jiggle in their local space. I believe even alias said that jiggle is good for fat tissue, because fat doesn't have a coherent structure like muscle fiber thus jiggles in the world space.

The second one is the surface that defines the muscle shape. Nurbs and Polys can tangle themselfs if you compress/twist the muscle to the limit.

Metaballs are the perfect solution, no matter how much tou compress a metaball muscle it will still have a nice/smooth surface.

As Simon said, and I'm not saying/implying anything else, wait for the 8.1 before you make a decision...

calin

Link to comment
Share on other sites

Some of the original papers written to do muscles were centered to using implict surfaces. Even some of the earlier code at ILM was based loosley.. If you look at PDI... again similar concepts. I think some of the second generation stuff at ILM was voxel based.. but again it's not a far leap from that to implict surfaces.

The hard part was for everyone to use them correctly. The common approach was to "ray" or fit a resulting surface from the implict one,but to do so correctly would require a lot of work to fill the volume. That's not really practical. You just need a smart way to use them.

As for the simluation stuff with characters.. it sounds great.. sure if you have all the disk space in the world and can 'bake' all the possible solutions to disk over night so an animator can use them the next day (cough PIXAR cough) simulations work great. IMHO. Velocity based stuff isn't as reliable to use in production as people think it is...the normal spring function itself will blow up given enough time... (vets of Houdini know what I mean). the JiggleCHOP is a bit interesting as it is biphasic in nature...it will give you the control you need so it doesn't blow up.

the inflate as it stands is an interesting idea.. not fully implemented IMHO, but we didn't choose to go that route on the wild in a large way. Like anything in this business you can get away with a lot of cheats to get a shot done. If it looks right/good then it is done.

Where was I?

-k

Link to comment
Share on other sites

@calin

thats sounds very interesting. it sounds like muscles in houdini?

i hope, the only thing is i have to decide in the next two weeks.

Maya seems good, houdini, too.

......very hard decision.

wich company's use houdini mainly at there work??

a52? DD? Framstore??? the mill?

these are companys, wich i wanna work for.

....

i am a maya user for more than 4 Years......

is houdini better.........

these are mainly very hard questions.....

i wanna a list with advantages and disadvantages.....

Link to comment
Share on other sites

as for which is better (a question that can't be answered) do a search here and on the Side Effects forums and you'll find lots of discussions about the comparative advantages and disadvantages of each program....

as for who uses Houdini:

http://www.odforce.net/wiki/index.php/CompaniesUsingHoudini

and if you're looking to get a job it really will NOT matter what program you use to do you reel...honestly it doesn't matter...if you show skill and ability you will be considered - that's it, knowing a particular program will only help a tiny bit...I can say this from personal experience of getting a job and not getting a job.

why do you have to decide in the next two weeks?

Link to comment
Share on other sites

i am a maya user for more than 4 Years......

If you've been a maya user for more than 4 years why do you have to throw something together last minute in order to apply for a position at these studios?

Link to comment
Share on other sites

i have made my decision. i stay with maya. till i got the job....or not.

Houdini is a very good tool in creating effect. But i think my experinces in Maya are so good and deep, that i stay with maya.

the reason to learn houdini was , that a friend told me, that some companys only wanna artist's wich can work with both, houdini and maya.

in germany, where i come from is that a very big problem, too.

If you are a dynamic artist or an artist in clothanimation or character anim...

you can't find a job, you have to do all.......Compositing, anim, relighting, and so on.........germany don't wanna get specialists

and so iam on the learning trip :D

But thats not my problem, i will go away from germany.....

to make VFX shots in commercial's in companies with good story's and with a good team.

because germany is not the good partner for these things.

the german's got no humor, and that sucks to work in germany.

.......

the next reason to learn houdini is, that i learned enough in maya, and i will look about the border, i will be a houdini master,too.

or maybe a good houdini artist.

i think its hard to switch in my head.........the valley between houdini and maya is so large that you have to make a choice for life???? :rolleyes:

i thanks to all.........

if somebody is interessted in some of my works....

go on: www.deathfall.com...........my nickname is mainframe123......there you can see some of my maya works...... and hopefully houdini works in the future.

http://www.deathfall.com/user.php?op=useri...me=Mainframe123

Link to comment
Share on other sites

that's the smart plan I think...

but no harm in continuing to learn Houdini...just knowing that it exists means you're ahead of the rest :)

and looking at some of the work you have posted at deathfall (why the hell is that site so damn slooooow...) I'd say you've done some very nice work...

Link to comment
Share on other sites

The second one is the surface that defines the muscle shape. Nurbs and Polys can tangle themselfs if you compress/twist the muscle to the limit.

Metaballs are the perfect solution, no matter how much tou compress a metaball muscle it will still have a nice/smooth surface.

As Simon said, and I'm not saying/implying anything else, wait for the 8.1 before you make a decision...

calin

26919[/snapback]

Calin, one thing I've always wondered about with using metaballs, even though they always provide a smooth surface you still have to tie the skin to them in some way, so if they are hitting extremes where Nurbs would tangle wouldn't a surface driven off them tangle too?

Of course you can use inflate but then there are two problems one you still need a skin to inflate onto them and that needs to behave like skin and stretch and compress correctly. Second problem is that by inflating you are saying that the metaballs need to accurately define the shape of the final skin. I suspect you have an answer to this, don't know if you can discuss it but i'm interested if you can.

Something I'm playing with at the moment is using a dummy nurbs surface as a skin that then drives your real skin via a wrap deformer (a sparse one that is pretty fast and efficient) Nurbs with high basis factors behave nice and smoothly and you'd have to push them pretty hard to make them tangle. They seem to quite nicely represent the way real skin flexes but with no overhead of a true sim. I can see it being easy to extend this with inflate to add muscle bulging and skin sliding. So in effect you end up with a universal rigged skin that can be quickly positioned inside any mesh and have it automatically captured. The nice thing being that you don't need a full set of muscles to perform a full capture you just use the ones you need to add an element of realism.

I'm also interested in whether the latest versions of the wire solver are capable of acting like muscle fibres, flexing, contracting etc. I still think ultimately that is going to be the best way to drive muscles or even directly moving the skin around (ala my flex system). I think I saw some things in the change logs that suggest the wire solver can now handle animated rest lengths, that would be key to creating a nice muscle fibre sim i think.

Roll on H8.1....... :D

Link to comment
Share on other sites

Calin, one thing I've always wondered about with using metaballs, even though they always provide a smooth surface you still have to tie the skin to them in some way, so if they are hitting extremes where Nurbs would tangle wouldn't a surface driven off them tangle too?

Of course you can use inflate but then there are two problems one you still need a skin to inflate onto them and that needs to behave like skin and stretch and compress correctly. Second problem is that by inflating you are saying that the metaballs need to accurately define the shape of the final skin. I suspect you have an answer to this, don't know if you can discuss it but i'm interested if you can.

Well, I can say something. By it's nature the meta capture will give you a more localized capture which means you'll have more control over your final mesh so you don't have tangled skin. The inflate will work on the unified/merged muscle surface so yes you'll need to build the metaball surface to some level of detail but on the other hand it will be much faster than any nurbs/poly surface in wrap mode. Wrap deformers can be realy slow if you have a heavy mesh and you can't have fine control over sliding and other skin attributes. I think that's about it, when comes to how much I can say about our solution at the moment... more to come in the future

Something I'm playing with at the moment is using a dummy nurbs surface as a skin that then drives your real skin via a wrap deformer (a sparse one that is pretty fast and efficient) Nurbs with high basis factors behave nice and smoothly and you'd have to push them pretty hard to make them tangle. They seem to quite nicely represent the way real skin flexes but with no overhead of a true sim. I can see it being easy to extend this with inflate to add muscle bulging and skin sliding. So in effect you end up with a universal rigged skin that can be quickly positioned inside any mesh and have it automatically captured. The nice thing being that you don't need a full set of muscles to perform a full capture you just use the ones you need to add an element of realism.

Nice, I would love to see the result...

I'm planning to build some muscle groups as digital assets (eg, biceps, trapezius, pectoralis) so in the end it should be pretty easy to set up.

I'm also interested in whether the latest versions of the wire solver are capable of acting like muscle fibres, flexing, contracting etc. I still think ultimately that is going to be the best way to drive muscles or even directly moving the skin around (ala my flex system). I think I saw some things in the change logs that suggest the wire solver can now handle animated rest lengths, that would be key to creating a nice muscle fibre sim i think.

Roll on H8.1.......  :D

26988[/snapback]

not sure at the moment

take care

calin

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...