WIPpa DoODLE (Tinman Revisited)

[Rodney]

This is more of a spline doodle than a WiP.

I started splining without any goal in mind and the next thing I knew I was modeling Tin Woodman.

 

For the animation portion (if you can really call it that) I was focused on two primary things; what expression could I make with solid shapes that don't articulate (i.e. how much squash and stretch could I get with shapes that don't change) and could I get a part of his neck design to double as a hint of lower teeth when his mouth opened to a full extent.

 

The basic splining followed an attempt to distill Tin Woodman down to the fewest lines possible for easy of drawing (primarily for storyboarding) that I never ended up using except for my own entertainment.

 

Added: This model started having enough 5 point patches that I found myself wanting a shortcut key to activate that so... I've assigned 'Control 5' as my new shortcut keys for creating a 5 point patch. This is often invoked immediately after the period key is hit twice to allow for those pesky 5 pointers that don't want to activate.

tinman revisited.mov

[robcat2075]

That's a great face, Rodney!

[robcat2075]

I really like the economy of that. You could slide the funnel/hat forward for downward eyebrows

 

and slide it back for upward eyebrows

And then rig the actual brows so you can scrunch them along with that motion.

 

You'd have a dialog ready face.

[Rodney]

Thanks Robert!

I realized pretty quickly that I was going to need some eyebrows to really sell any emotion but haven't moved to that stage.

As much as I've been thinking about faces of late it's almost enough to make me suspect I'm trying to build up my courage to get into rigging of a face.

Yikes... am I really thinking that???

Could I be laying aside my laziness and moving toward those things that I really need to have in my toolset in order to be successful???

 

I have had the itch to work with some dialogue but not just for the dialogue's sake... more to better understand how to rig for expressive dialogue as well as to gain the experience.

 

This is the second tinman face I doodled today. The first one having boolean eyesockets cut out of the face.

I then realized that would never work as good as actual geometry for the eyesockets so shook off my laziness and dove into creating the 'real' thing.

 

I've attached the actual model to demonstrate just how lazy I am.

Please excuse the bones... as they are just quickly thrown into place.

This may be the one with normals of the main head facing the wrong way...

Tinman000.mdl

[Rodney]

Right after Tinman I doodled Froghead.

Um... I have no idea why I splined him either.

I think I just wanted to model a mouth with teeth.

 

This uses a combination of A:M's sharpen and blur GPU post effects.

I kinda like that color outline that sharpen creates.

[robcat2075]

I took a quick look at Tinman. i guess it will need a few more spline so the five-pointers aren't part of such tight turns. They look OK in shaded mode but in final mode the five-pointers are broken up.

[Rodney]

They look OK in shaded mode but in final mode the five-pointers are broken up.

 

Yes, that's what appears to be giving the render the discoloration (black detail) around the eyes.

Although some of that is mixed in with the reflection.

[robcat2075]

There are quite a few patches facing inward, but even after I flip those the 5-pointers are getting stressed out.

[Rodney]

even after I flip those the 5-pointers are getting stressed out.

 

Stressed Out 5 Point Patches... it's a feature.

 

... like on the tip of Tinman's nose.

[robcat2075]

Looking at it a bit more... it seemed to me that the 5-pointers shouldn't be having so many problems. i don't remember having that much creasing to them even on highly curved ones.

 

When i delete all the groups in the model they get much smoother. So there's probably some sort of conflict in a group. I'm not sure where but it's a clue.

[Rodney]

I belive one of those groups has 'Average Normals' set to on.

My memory tells me that it's best if that group is the first one in the entire listing of groups... and it is not.

Maybe that relates to what we are seeing?

 

When I toggle that Average Normals setting to 'Off' a lot of artifacts go bye-bye.

 

For what it's worth I see a lot of named groups are in the wrong order too.

For example the black surfaced groups for the eye sockets don't appear because they are overwritten later by the 'reflect' group.

 

This is another area where learning occurs... if these groups were all named it'd be much easier to know at a glance what their role is suppose to be.

 

 

Added: I note that the eye sockets should have another ring just inside to limit the extensive curvature at that location. Peaking those splines that enter into the eye socket removes a lot of unwanted orientation.

[Rodney]

Here's a slightly updated Tinman.

Changes include:

 

- Got rid of the 'Average Normals'

- Named most of the groups

- Modeled a slightly more proper jaw - the previous one wouldn't work mechanically - it was just there for looks

The new one still needs a lot of work

- Eye Sockets adjusted

- A few other things I've forgotten

 

To Do:

- Place bone to be better aligned to proper pivot points

- Figure out why reflectivity shows only when normals are inverted and disappears when pointing out (this is odd)

- Eyebrows

- Termination of neck and refinement of (inner) mouth - at that point I obviously just stopped modeling

- Probably need a spline ring on the outside (facial surface) of the eye sockets - Possibly even a separate piece to suggest joining of real metal

 

Perhaps this is a WIP after all... so all suggestions for improvement appreciated!

 

For folks that wonder where this guy might fit into Oz continuity, I'd say this guy is a part of the Tin Army (from a very short and short lived story by Baum).

As such we'll probably have to exchange the tin hat for some kind of helmet.

Tinman004.mdl

[detbear]

Rodney,

 

Very Cool. I think I like that style even better than the original Tin Man Version.

 

Nothing against the old version of course.

[Rodney]

With regard to the eyes/eyesockets perhaps a fitting something like this... (maybe even with a few tiny rivets)...

 

Edit: Thanks William!

The Tinman I'd really like to create would be short/tiny... cute even. But I have no plans to go there anytime soon.

[Rodney]

Now with patent pending rivets...

 

 

[Rodney]

Thought I'd try to get the nose and ears mirroring the general style of the eyes...

 

Tinman v4a is attached.

Edit: Fixed the jaw/jaw bone and uploaded again.

Added upper teeth as well.

Tinman004a.mdl

[robcat2075]

I'm surprised the 5-point rings could be made into a patch. It shouldn't be able to do that with one spline.

[Rodney]

It shouldn't be able to do that with one spline.

 

I've been trying to convince folks otherwise for years... many years.

I believe I first stubbled upon the capability when I was trying to close shapes for use with boolean cutters.

The secret (usually) is to use the comma key to select the entire spline rather than another method to grab the whole thing.

ALTHOUGH in current releases we often don't even need to do that.

 

Here's a shot of some quickly splined/copy pasted flower petals... all are created via 5 point patches.

[Rodney]

Just when I thought I knew the limits of 5 point patches along came Malo and demonstrated that I didn't.

Here's a single spline with 4 - 5 point patches.

 

 

At the outset I mostly used 5 pointers as a quick and easy way to cap the ends of stuff.

Lathe things with 5 sections and you can always turn the ends into 5 point patches.

This is especially useful when the objects are small and the ends don't need anything fancy... just need to be capped.

[Rodney]

Here's a cyclinder lathed with 5 cross sections and capped to make it look like it is filled with water.

For objects that just need to be tossed in the background this approach can lower spline/patch counts considerably because not only are we limiting the number of splines/patches but we are reutilizing a part of them by capping them as 5 point patches.

 

Worth noting: a downside of lathing with 5 cross sections is that the center of the model isn't always where you think it might be (i.e .when looking at the object from the cardinal directions; front, left, right and back.

[Rodney]

Copy pasting almost always retains the 5 point caps/patches...

Useful for speedily duplicating objects that are closed/capped.

That way we don't have to go in to each object and close or cap.

I mention this because this wasn't always the case in earlier releases.

[Rodney]

Decals work well with 5 point caps.

Patch images... not so much.

 

 

[Rodney]

Possibly related.

 

I think this is the guy who dares to wake Tinman up from his eternal slumber...

[robcat2075]

Is that Lightbulb King?

[Rodney]

Is that Lightbulb King?

 

 

If he isn't he sure could be!

 

His motto: "I may not be the brightest bulb in the box but some day I might be."

[robcat2075]

If I were to make an informal list of face features that one right rig...

(starting with the most important)

 

-mouth open/close

(this is all Muppets have and they do quite a bit with it)

 

-eyeballs pointable

-eyelids movable/blinkable

(to do something besides the straight ahead stare)

 

 

-eyebrows poseable
(there's probably more expression potential in the eyebrows than there is in the mouth)

 

 

-mouth shapeable
(this is what everyone thinks they need for dialog, but I think people obsess about it too much)

 

After that there is a long list of lesser elements like tongues, teeth, cheekbone scrunches, nose flares...

[Rodney]

Thanks Robert your thoughts on dialogue align with what I think is important (priority-wise).

I keep thinking that where it comes to mouth shapes I'm mostly insterested in a japanese phonetic approach to shaping (primary of the open/close variety). Namely:

 

A (Ah - as in 'Ah, so that's how it's gonna be')

I (E - as in the letter E)

U (Ooo - long drawn out)

E (Eh - as in 'Eh? What was that?)
O (Oh - as in 'Oh, Oh, that's gonna leave a mark')

 

Those five shapes are the primary open/close variations.

The accents then are the consonents

Ka

Ki

Ku

Ke

Ko

 

Sa

Shi

Su

Se

So

 

Ma

Mi

Mu

Me

Mo

 

Ra

Ri

Ru

Re

Ro

 

Ha

Hi

Fu

He

Ho

 

Pa

Pi

Pu

Pe

Po

 

Etc., Etc.

 

M and N are mmmmostly closed (M usually being the initiator and N being the closer)

 

I'd say a lot of personality shots can be pushed by moving the jaw slightly to the side (ala a person who speaks largely out of one side or the other of his mouth).

 

As you well know, the joy of dialogue isn't even in the facial movement but in the accent of the body itself, with hand gestures and poses that convey something other than what is being conveyed in the dialogue.

 

Too often we see dialogue heavy pieces that explain (or reexplain) what the visuals have already sold to the audience.

Whereas the dialogue would be better served by revealing through words what is in the mind of the character that the visuals don't show.

For instance, a disinterested character that is pretending to pay attention.

Or perhaps better yet... as a foreshadowing of something that will be returned to later as we learn more about the character.

 

As you say though, I think the simple open/close of the mouth can serve the job best in many cases.

Then it's mostly a matter of tying the facial poses into the intended emotion of the character.

 

This is even further demonstrated by the many shots that have a character talking while the camera is fully focused on the reaction of another person who is (suppose to be) hearing the dialogue.

In those instances not only does the sound hit its cue but we get to see the reaction of what another character perceives concerning the dialogue.

It is through that secondary character that we (in the audience) are able to assimilate the true intent of the purpose/existence of the dialogue.

Time after time after time we see that most dialogue is unnecessary and yet scripts are still packed full of the stuff.

 

I see two primary purposes to dialogue; the first is to intentionally lead someone astray (regarding the plot); i.e. to believe something other than that which needs to be reserved or withheld until later in the plot. The second is to clarify and or resolve such ambiguity so that misunderstandings do not happen. Of course, some dialogue intentionally leads us to believe that we, along with the characters, don't have a clue what has happened (i.e. dangling plots).

 

Added: I do think that something most cg characters lack is ability to really exaggerate mouth shapes. We really need to be able to show those toothy grins, puckered lips and such to gain enough exaggeration to push those personalities. Of course this is mostly for characters whose performance will be highly energetic. One who isn't energetic might have very little mouth movement consisting of mumbling to themselves.

[John Bigboote]

Great stuff! Very informative... I love the show-n-tell, our forum needs more imagery.

[Rodney]

I love the show-n-tell, our forum needs more imagery.

 

 

Thanks Matt.

 

I agree. And that's why we have a WIP section in the forum in the first place.

It's kind of like posting in a personal blog... but more interactive.