Building the rig: eyeball control

First feature test, roughed-out head working towards the tough Spitfire pilot. Soon will be able to have him tracking the sky. Quite complex though. Which is nice.

Mrs Ming suggested a good one. The guy looking ahead the thousand yards and then the sound track has high heels walking past. Can do something with that soon :slight_smile:

Grains of rice correct that man :slight_smile:

Ming

https://s117.photobucket.com/user/Mington/library?action=view&current=EyeballsControllerTest.flv

nice but right eye? :wink:

By comparison here’s the default XSI ‘Man’ object fully rigged. I’m just animating the LookAt controller. He’s a bit elfin, a bit rabid. Our guys will be chunky but lovable :slight_smile:

https://s117.photobucket.com/user/Mington/library?action=view&current=XSIdefaultguy.flv

Bearing in mind that there is the HL2 content creator addon for XSI… ValveSource on the menu bar, practically drag and drop objects from XSI into a HL2 modded game, which is part of The Plan

The Cinematic Mod for Half-Life 2 is getting improved further: The upcoming version 9 includes better performance on 32 bit systems and special shaders for a more detailed outcome of character skins.

http://www.pcgameshardware.com/aid,658914/Imagegallery/Fakefactory_Cinematic_Mod_9_first_screenshots_of_the_new_version_with_extraordinary_skin_textures/
Look at the ‘original’ version of this shot there, crikey

Ming

It’s a feature test Wildcat. Can it be done given what we know? Everything else is of no consequence

He’s lucky he’s got the one eye Admiral :slight_smile:

Ming

Your eyelid closing “wink” sequence is probably not the way to go Ming. Now character animation is really not my “thing” but i believe what is usualy used for facial animations are PLA morphs.

PLA morphs are “point level animation”. The eyelid is modeled in open, closed and expresive positions using the same vertex construction and set as morph targets. The facial expresions are then bound to a controller object which is then animated.

This alows for differing eye/facial morphs to be blended. The art being in creating the individual morphs then controling them via a controller object.

It makes life a lot simpler in the long run, especialy if the face is going to used extensivly and speach sync is to be attempted.

HTH m8

Sporran

XSI has it covered no worries Sporran thanks mate, you’re describing what I’m doing right now but Softimage obviously have different handles for stuff. I told you about Shape Manager (morphing, that first-smile clip) earlier. Morphing is quick to set up but I’m working on vertex animation via bones to learn the bones’ ropes I’m always killing two birds with one stone. I’m working too intensively on four things at once (as I’ve learned you 3D guys absolutely need to, to get up to speed) at the moment to spare time to show more finished things. Focusing on that late afternoon scene looking south. Head modelling in parallel with animation, working them up together

Yes exactly what you say mate, that’s where I’m up to on this stuff. Building the controllers. Can be 3D objects in the workspace, very exciting to have high-level controls. But I must focus so that one is me just controlling the eyeball’s constraint manually. He’ll be following things later. And focusing on an enemy plane, then you’re swept into the back of his retina with the electrical activity and a red recognition light coming on. That’s a bandit Boothroyd :slight_smile:

That was a test that eyes can be controlled- that I’ve learned how it can be done. Loads of linking parameters to control objects and so on but it’ll be a while before he starts looking interesting. I’ve got seven days as usual :slight_smile:

You should see this XSI kit mate I’m working up to something quite special in the event I don’t fall under an ambulance :slight_smile:

There’s also in this version that thing where you render different passes and then er, recombine them offline to create the final render, in post-processing. It’ll come to me :slight_smile:

Found out why some clips are large. It’s the ones with the ‘Shaded With Mesh Visible’ for some reason, compression thing. They’ll need to be stills. Sort of defeats the object though :slight_smile:

Ming

that thing where you render different passes and then er, recombine them offline to create the final render, in post-processing. It’ll come to me

Compositing :slight_smile:

This ICE stuff is looking good and naturalistic…

Ok back to the head modelling, couldn’t fix the head without knowing how the vertex animation would pan out. Chickens, eggs but appears to be feasible. Fusible maybe :slight_smile:

These full-on bald heads are a bit weird, he’ll have a wig next time or my name’s not Terry Wogan

Ming

Not morphed this eyeball test Sporran and I’ll be using a lattice transform to give the airman normal-shaped oval eyeballs. I’m trying out the various functions for simulating/controlling eyeball functions. Should have him looking around suspiciously shortly :slight_smile:

The eyeball (in the clip, the image below has just placeholder ping-pong balls) is a surface rather than a polygon object, with the pupil-object constrained to the surface. Learning all that complex ‘Link With’ stuff, phew it’s mind-boggling but getting there

Now using a (deceptively) simple model for tidying up ready for clean control of eyeball functions via an animation control panel with sliders. Concentrating on the eyeball control functions as that’s where the most expressive bits will need to be animated

Ming

https://s117.photobucket.com/user/Mington/library?action=view&current=EyeballFunctions_1.flv

Bit more work on him, looking like a death mask which is encouraging :slight_smile:

PLA morphs are “point level animation”. The eyelid is modeled in open, closed and expresive positions using the same vertex construction and set as morph targets. The facial expresions are then bound to a controller object which is then animated.

This alows for differing eye/facial morphs to be blended. The art being in creating the individual morphs then controling them via a controller object.

It makes life a lot simpler in the long run, especialy if the face is going to used extensivly and speach sync is to be attempted.

That’s exactly what I’m doing Sporran - now it’s a couple of days later when I actually know what I’ve been doing as I’ve been following(stumbling through) tutorials :slight_smile: - and you nailed it again yes, it’s for re-useable content that I’m concentrating on working the whole lot up together rather than getting bogged down in the modelling. Makes things more interesting too, switching to a bit of modeling then into studying control objects and so on

Send me on your German armour and I’ll get them pounding around the place in a HL2 mod. Not this year though :slight_smile:

Ming

are as my eyes: tired!