Tuckertown and other Stories

[Rodney]

Welp... found that one!

 

Now I remember why I set that entry aside and didn't dive into the last 10% polishing.

I really wanted to get a 'bursting through your screen' effect but got off to a bad start on achieving that.

The reason for misplacing the file? I cleaned off my desktop and pushed the files it into a temporary folder named "AMData".

 

These two guys are related to the story with the (greenish) cat.

They are time and space traveling.

[Rodney]

This may be sacrilege so 3D purists... please forgive.

(I used pure A:M for my contest entries. I don't think I even used any externally generated decals! (see my previous post also for a look at a render out of A:M that is purely A:M with nothing else in the mix)

 

I confess that when I see the lines between 2D and 3D blur I feel a tinge of excitement.

Both 2D and 3D are great on their own but when combined together seamlessly they can be pretty special.

 

Here is an example of a simple mesh created in A:M being used in a 2D animation program.

For those of you that have been following my blog you'll recognize the program.

 

Of interest though is that of determining what parts are 2D and what are 3D.

 

It's a simple animation.

Does anyone care to guess what is what?

 

You'll likely have to click on the image to view the animation.

 

 

 

Aside: I'm currently troubleshooting an issue with the rendering of gif transparency so that is why the animated gif has a white background.

[fae_alba]

the eyeball is 3d.

[Rodney]

the eyeball is 3d.

 

Nope. 2d shapes with shaded fill.

 

 

I'll try not to leave you hanging too long but will wait to see if anyone else cares to guess.

There isn't a whole lot more to guess.

[fae_alba]

Then that leaves the bit around the eye. But then I'd have to ask what's the advantage? Is it more a means to have a character that is part 2D and part 3D?

[Rodney]

Paul,

I wrote a few paragraphs outlining some of the process and potential but in realizing that words won't be adequate I resolve to put something together that can demonstrate the benefits.

There is a lot going on in that image of an eye but much of it is a bit esoteric and none of it really relates to that of animating an eye... other than that was the shape that I decided to quickly put together.

 

If you've got a few shots of your 'Papa Bear' project you'd be willing to share perhaps I could put together a small demo that outlines some useful processes related to this.

Barring that I'll eventually get there but I don't have any specific projects I'm working on so that's why you see things like strange eyes and such. I often decide to test a process and use whatever I think of at the time that gets me there the quickest.

[fae_alba]

Paul,

I wrote a few paragraphs outlining some of the process and potential but in realizing that words won't be adequate I resolve to put something together that can demonstrate the benefits.

There is a lot going on in that image of an eye but much of it is a bit esoteric and none of it really relates to that of animating an eye... other than that was the shape that I decided to quickly put together.

 

If you've got a few shots of your 'Papa Bear' project you'd be willing to share perhaps I could put together a small demo that outlines some useful processes related to this.

Barring that I'll eventually get there but I don't have any specific projects I'm working on so that's why you see things like strange eyes and such. I often decide to test a process and use whatever I think of at the time that gets me there the quickest.

Rodney,

 

Papa Bear is always willing to be a guinea pig! Do you want a still, a model, or what?

[Rodney]

Rodney,

Papa Bear is always willing to be a guinea pig! Do you want a still, a model, or what?

 

When I was typing I envisioned a short sequence that includes a backgrond, foreground and middle ground with the various elements involved (all the objects/models) there.

This could be a scene that doesn't have any characters in it such as the opening scene. It could also be something that won't appear in the short itself (think of the various film credits that take the audience back through scenes from the film but in a different style than the film itself such as a 3D animated film that has hand drawn scenes recreated for the credits. Barring that the simplest thing would be a model... any type will do to test out the possibilities.

[fae_alba]

 

Rodney,

Papa Bear is always willing to be a guinea pig! Do you want a still, a model, or what?

 

When I was typing I envisioned a short sequence that includes a backgrond, foreground and middle ground with the various elements involved (all the objects/models) there.

This could be a scene that doesn't have any characters in it such as the opening scene. It could also be something that won't appear in the short itself (think of the various film credits that take the audience back through scenes from the film but in a different style than the film itself such as a 3D animated film that has hand drawn scenes recreated for the credits. Barring that the simplest thing would be a model... any type will do to test out the possibilities.

 

 

how's this...

 

[Rodney]

I'd say the character himself would be great. (Somewhere I think I have an earlier version of that bear)

Are those buildings in the background models?

If they are models, I'd say 'yes' to them as well.

If it's an image backdrop... probably best to replace it (for our purposes here).

 

The golden rule I'm operating from at present is:

- Use A:M only for first order resources (this encompasses the 3D 'purist' view but then goes one step further... if the decal that is used isn't created in A:M via materials or rendered models it's not first order so better left out at the first stage). This relates to efficiency that only exist in the world of A:M and important to consider if we are to fully leverage A:M.

- I'm still working on the second and subsequent orders which relates to why I posted my test here but it moves us more fully into the world of imports, exports conversions/translations and compositions.

[fae_alba]

I'd say the character himself would be great. (Somewhere I think I have an earlier version of that bear)

Are those buildings in the background models?

If they are models, I'd say 'yes' to them as well.

If it's an image backdrop... probably best to replace it (for our purposes here).

 

The golden rule I'm operating from at present is:

- Use A:M only for first order resources (this encompasses the 3D 'purist' view but then goes one step further... if the decal that is used isn't created in A:M via materials or rendered models it's not first order so better left out at the first stage). This relates to efficiency that only exist in the world of A:M and important to consider if we are to fully leverage A:M.

- I'm still working on the second and subsequent orders which relates to why I posted my test here but it moves us more fully into the world of imports, exports conversions/translations and compositions.

 

Yeah, the buildings are just a backdrop. How about this one? Each object is a separate model and I can send you an archive with everything in it.

 

[Rodney]

That'd be great!

Hopefully you still have my email address.

[fae_alba]

just sent an archive off to you...enjoy!

[Rodney]

Thanks Paul. Got it.

[fae_alba]

Thanks Paul. Got it.

 

Lookin forward to seeing what magic you work on it Rodney

[Rodney]

I haven't had much time to investigate the possibilities but I did conduct a quick inventory of what is in your project file.

I'm a firm believer that each project and scene brings to it unique considerations that when considered in part inform the larger whole of the complete project.

There is also the theme and general message to consider... .what am I trying to convey... what is the point of the project.

Those are just some of the many things I want to consider as I dive into your project.

[fae_alba]

I haven't had much time to investigate the possibilities but I did conduct a quick inventory of what is in your project file.

I'm a firm believer that each project and scene brings to it unique considerations that when considered in part inform the larger whole of the complete project.

There is also the theme and general message to consider... .what am I trying to convey... what is the point of the project.

Those are just some of the many things I want to consider as I dive into your project.

 

Rodney I was striving for a film noir detective movie kind of vibe. It will be interesting to see if you can pull it off!

[Rodney]

I haven't forgotten about this Paul... still pondering the possibilities and off being sidetracked.

I'm pretty sure it eventually leads back to here.

 

I was striving for a film noir detective movie kind of vibe.

 

I can definitely see that!

[fae_alba]

No worries... I've been busy playing dodge-em with Hurricane Matthew. Was sent to Miami on Tuesday for a client meeting. Upon landing I got a text saying the client is delaying due to the impending storm (but to come in on Wednesday as scheduled and we'll "keep an I on the storm"). I said "not!" scrambled and got a flight out yesterday at 3am.

[Rodney]

I said "not!" scrambled and got a flight out yesterday at 3am.

 

Sounds like a good decision to me.

[Wildsided]

 

I said "not!" scrambled and got a flight out yesterday at 3am.

 

Sounds like a good decision to me.

 

seconded

[fae_alba]

 

I said "not!" scrambled and got a flight out yesterday at 3am.

 

Sounds like a good decision to me.

 

 

In all likelihood I'll be sent back down sometime next week and I won't find out till Sunday. makes planning a tad bit difficult.

[Rodney]

Random (Baby) Groot...

 

I make no claim this looks much like Groot as I was doodling from memory and a general feeling.

One of these days I'm going to start actually refining my doodles with reference but that can take a little of the fun out of it and the random doodles are mostly just to keep the muscle memory of modeling and to explore how to quickly and more optimally put splines and patches together.

 

 

 

 

[Rodney]

And here's the reason we should use reference kids...

 

In the attached, I can see many areas that would improve my doodle, plus it and get it moving toward the goal of 'being Groot'.

Those eyes for one. That would have been easy to change.

I like the contrast of the character against the blue background too!

 

P.S. My groot is known to say, "I am not Groot".

[Rodney]

I doubt I'll pursue this much more but I figure I'd best bring bg a little closer to the reference...

[Rodney]

Random crowdtest...

 

It might seem overly ambitious but I can't help but think we can achieve Zootopia level quality of crowd movement with pretty basic setup and straightforward workflow.

 

This setup and execution only takes a few minutes.

Swapping the proxies out for articulated characters with walk cycles would take considerably more time in order to suggest proper detail and personality.

No particles used.

[Rodney]

Random island created with the Terrain Wizard....

 

(The secret sauce in the vegetation and water is surface roughness)

[Rodney]

Random fox-like character's head...

 

This was an experiment in modeling from a simple grid* and working toward as simple a head as possible.

I won't say it was an entire success because the model didn't stay as simple as I had hoped but it did exercise my modeling muscles a little.

 

I've attached an image with three of the initial grid progressions that (eventually) was modified into the head.

It should be noted that deleting the splines/CPs in the mouth and eyes was not strickly required... it was based on prior experience.

Leaving those in the grid would very likely have saved me some struggling later on when my brain had to shift from 'stitching' details into place to 'extruding' forms in 3D space. My thought now is that it would be better to stay in the 'stitching model' as long as possible before distorting the grid into shape and might have the advantage of removing the requirement for extrusions entirely. Keeping the patches flat would also be useful when considering additional detail to be added via decals and texturing.

 

 

*Mark Largento mentioned this approach the other day and we must have been inspired by something or someone similar because I had just experimented with the grid-modeling approach the day prior. Coincidence? I don't think so but the source of that approach is a mystery to me short of playing with features found in v19 Beta... so perhaps that experimentation in the beta led to that form of experimentation.

[fae_alba]

That's an interesting character. For some reason I can't envision how you got from a flat plain to the 3D head. I always have a problem with drawing in 3d space until I have a nearly completed model. That's why I start with existing models and work from that.

 

As an aside, it'd be interesting to see what kind of rig you'd come up with for the mouth/snout. That's one big piece that I haven't been able to do nicely with Papa Bear.

[Rodney]

it'd be interesting to see what kind of rig you'd come up with for the mouth/snout. That's one big piece that I haven't been able to do nicely with Papa Bear.

 

I've yet to come up with a rigging approach I like for a fully articulate... and highly expressive... exaggeration-ready... mouth but...

 

As far as I can ascertain there are two basic needs for rigging of a mouth. The first is the jaw (rotation and lowering) to get at the basic shape of an open mouth that can stretch while still maintaining a since of the solid jaw underneath. The second is a ring of bones around the mouth to provide the shapes necessary for lipsync/dialogue.

 

The most successful tests I've come up with only have the jaw bone and the mouth shapes are manually animated. This matches the general approach of lipsync with phonemes ala the Dopesheet/poses.

 

Two areas of concern seem to be most prevalent when modeling the lower face and the first is that the mouth needs to wrap around to the sides of the face at least enough to gain the ability to pose the mouth in front, side and (perhaps most importantly) three quarter views to aid in the viewer actually seeing the shapes.

 

My own goal is to make sure the mouth can be moved into extreme positions beyond simple dialogue shapes in order to target and really push expressive emotions. When drawing these extremes can be pretty easy to achieve but with a rigged character the animator may find they've hit a limit that they can't move past that they require to get the shot they want.

 

I've found a useful approach to modeling the mouth is to model it in an extreme open position first, to more easily have access to all the control points, then adjust the jaw/mouth/lips back into a proper position for a relaxed state. I can't say that I've achieved success with this yet but I've failed enough in the attempt to know what doesn't work so well.

 

I can see where Papa Bear's mouth might prove troublesome.

Like most articulate faces I would think the snout would be maintained with the upper skull (and move very little) and the mouth follow the jaw.as it rotates and moves. The skinning/mesh between skull and jaw then needs to be able to accommodate the resulting squash and stretch.when the jaw's movement forces changes of shape.

 

 

I can't envision how you got from a flat plain to the 3D head.

 

As with most things I always think I'll record a session and share it but I haven't quite got to that point yet.

 

I will say that it might be well worth the time to model the skull area and the jaw area separately and then later stitch them together.

Thus far I've mostly been modeling them together.

 

To get from 2D plane to 3D shape I generally start with a 3x3x3 distortion cage, grab the middle four CPs of the cage and move them forward (from side view) and then start to adjust and fill in shapes and details. There is always grunt work to do... fill in 5 point patches... add additional spline rings around eyes, nose and mouth... etc.

[Rodney]

Here's a variation on the theme...

 

In this case I started with the four 'holes' of the face*; 2 eyes, nose and mouth and modeled them toward the basic grid shape. Then distorted them from that 2D plane into 3D space.

In this one the mouth doesn't quite get enough splines flowing into it to get at proper shaping.

Unlike most of the grid-models I've made I modeled the eyes as connected parts of this face. Usually the eyes work better as spherical objects that can be rotated separately.

 

 

*Technically I guess we could say there are seven 'holes' in the head (if we don't count nostrils) when we include the two ears and the neck.

 

Added the initial grid. Those with an eye for detail may note that I've reformed the splines that move down just to right/left of mouth so that they wrap around to follow the contour of the jaw and connect.

[Rodney]

Here's a quick try on a bear shaped face... with spherical eyes.

 

(The grid shows mid way stage right after thinking 'I should add some nostrils')

 

 

Yes, all these characters need teeth.

[Rodney]

Random teeth...

 

 

[itsjustme]

That's an interesting approach, Rodney! There are some splines I would re-route once you got past the initial stage, but it seems to be a great way to start.

 

Smartskin and Poses could help quite a bit in rigging a snout.

[Rodney]

There are some splines I would re-route once you got past the initial stage

 

Most definitely a requirement. The initial grid is never left unmodified... at least not for me... I'm not that good at anticipating where the curvature will work best in 2D so that it will fall into proper place when properly distorted... not yet!

 

One of the neat things about the grid-modeling approach; the ability to explore different topologies quickly and reroute splines as necessary first in the 2D plane and then later again in 3D as we discover something doesn't quite work. The knowledge gained is then applied to the next attempt... and the next... and the next... and perhaps best of all, when going back to standard modeling sans grid the brain (usually) remembers the lessons learned and the modeling comes together without as much guess work.

 

I plan to use the grid modeling approach with more mechanical shapes as well but I've been having fun with modeling heads at the moment.

I assume mechanical modeling will lend itself readily to the grid structure although time will have to validate that.

 

Aside: One useful aspect of grid-modeling (in general) is how patches tend to start as roughly the same size.

This is ideal in many cases and many of the models I've deemed as having ideal topology tend to keep patch sizes very close to the same which tends to maximize spline/patch smoothness..

In modifying the initial grid I tend to lose consistency in patch size and need to remind myself to do my best to move back toward that.

 

 

 

I did try some cube-modeling with use of v19's 'remove internal patch' feature but the primary issue there is one of spline continuity. There is no way A:M can anticipate what the best flow for a spline will be in every case so that is usually best left to the modeler to determine. Like the grid-modeling method, cube-modeling can give us an initial starting point but unless it's used only as a proxy (not likely to be seen up close) it will require work.

[Rodney]

Taking a slight break from modeling faces (but not quite escaping from grids), here's a terrain generated using Landscaper material (color), Terrain Wizard (geometry) and something I'm calling A:M Pixelpainter (to generate the image used by Terrain Wizard). In short, everything created with A:M. A:M pixelpainter is a foray into gridpainting via groups... on odd concept that worked much better than I had anticipated.

 

I saw an old topic where Andy Whitlock mentioned he was willing to share the Terrain Wizard code so that others could work toward other terrain plugins. I'm guessing he'd still be willing to do that. Of interest to me is the Terrain Wizard's ability to paint inside a plugin window inside A:M... that certainly has potential for other applications.

 

At any rate... a random terrain...

[Rodney]

And another pixelgridpainted terrain.

This one leverages the seldom used Composite feature as the decal image applied to the displaced terrain (which was captured using A:M's Snapshot feature because I hadn't used that in quite awhile).

 

Why use Composite instead of a typical image as a decal?

Composite introduces a non-destructive workflow because any change made in the composite can then flow to the decal.

This leaves original images contained in the composite in their original state but changes the output via 'live' post effects.

If a decal image is used, that image must be modified/replaced.

[Rodney]

I can't quite remember who it was at Hash Inc (I think it was Heath) but I recall exchanging a few emails with one of the programmers where I exclaimed, "We can write our own programs with A:M."

The response I got was one of those deadly silences that implied a straightjacket was waiting in the wings for me ala, "Yeah, sure thing Rodney.".

 

Grid paint isn't really an example of what I was talking about then but it certainly comes close.

 

Basic steps:

Create grid of desired dimensions

Create group for colors needed

Use lasso tool, patch selection tool or other to select parts of the grid.

Want to clear the 'canvas' and start again? Delete all of the color groups except the base color group.

 

Additonal features

Right click and create a material of the current color (apply that colored material as needed)

Copy paste painted geometry into 3D space to achieve effects with depth.

For more interesting colors/fills change the material to something more complex.

For access to better results and more tools consider gridpainting in a Chor window.

 

So yeah, if you get a little bored... create your own mini programs with A:M.

 

 

Added: Example of basic use in Chor window

[fae_alba]

 

 

I've found a useful approach to modeling the mouth is to model it in an extreme open position first, to more easily have access to all the control points, then adjust the jaw/mouth/lips back into a proper position for a relaxed state. I can't say that I've achieved success with this yet but I've failed enough in the attempt to know what doesn't work so well.

 

I can see where Papa Bear's mouth might prove troublesome.

Like most articulate faces I would think the snout would be maintained with the upper skull (and move very little) and the mouth follow the jaw.as it rotates and moves. The skinning/mesh between skull and jaw then needs to be able to accommodate the resulting squash and stretch.when the jaw's movement forces changes of shape.

 

 

 

The below images of Disney's Humphrey the Bear are along the lines of the sort of expressions that Papa Bear needs to emulate. For right now I don't see him speaking, but rather like Humphrey pantomiming

 

[Rodney]

Well, I wrote a post in response but in the process of pursuing a quick grid-theory I seem to have browsed away from it.

Ah well... most of it was theoretical outside of my initial 'Nice reference!'

 

A few things I recall.

- We have to allow for artist differences in that reference.

- Our exploration can benefit from the fact that eyes and body don't have to be explored... but (especially given what you say about Papa Bear not talking) the eyes and body will most certainly have to be added back into the equation

- The snout in the reference appears rigid, steady and independent of the mouth. It's a pointer... seeking... searching... for the thing of interest... smell of food...etc. The primary part that I would think would be animated would be the end of the snout/nose with the nostrils... assuming that the present story calls for that, which if I recall correctly it does not.

 

 

Here's a quick take on what I think might work... still using the grid-method because I haven't tired of it yet.

I was tempted to include the tongue in the grid but then thought better of it.

 

Now to do a few rig tests.

 

[Rodney]

Exploration.

 

Snout as child of Head.

Jaw as child of Snout.

 

Orient Head.

Aim Snout.

Articulate Jaw.

Pose Lips.

 

This guy turned into a dog (literally and figuratively?) when I decided your character would likely need to whistle.

Because the shape looked more like a dog than a bear the whistle turned into a howl.

(Eye is notional because this thing looked strange without any eyes)

 

 

More splines will be required to get at more expressive shapes.

Less spline required in some places which makes me want to consider hooks (something I seldom use).

 

Added: Shot of bone movement.

[fae_alba]

Now that looks promising. The other place for expression is the upper lips/jowls. They need to have a flappy kind of vibe. But what you have there has quite a bit of expression in its own. My holy grail would be able to mimic that sheepish look that Humphrey often gets.

[Maniac]

very nice models

[Rodney]

More randomness... (I don't want to let a day go by without splining something in A:M)

 

This started as a quick doodle of making flat anime-style eyes.

The eyes needed a face...

The face needed a body...

The body needed an environment...

And this strange slugworm thingy appeared shortly thereafter.

 

In these explorations I try to include things I haven't done before or tried in a while. Those include:

Experiments with mixing of color, light and transparency.

Playing with Mist.

Bias tweaking. (well I always use that but usually not in negative values)

 

And no, I have no idea how this guy moves across the ground with a shell on the ground.

Some kind of magic drawing power?

 

I did say these were random!

[Rodney]

Now that looks promising. The other place for expression is the upper lips/jowls.

 

I put splines in the mouth to accommodate the first part (upper lip) I think the splines might be adequate for the jowls (although I hesitate to commit to that).

After staging the presentation showing the bones I regretted not using a slightly different angle that could account for more of the upper lip as it rests under the snout.

My thought here is that some key poses would need to be set to easily dial up those facial expressions that emote enough to capture that feeling of a mind inside the character (usually doing something slightly different than what their outer performance suggests). You definitely wouldn't want to manually animate splines/CPs every time you wanted to strike those poses.

 

I may start afresh and try to get the model looking a bit more like Humphrey/Papa Bear.

Alternately (additionally?), perhaps you can share a current version of Papa Bear and I can explore that. I think I have an early version of Papa Bear but you've surely changed him since that.

[Rodney]

Random surface bake tests...

[Rodney]

Q: What happens when we join four cubes together?

A: That depends on how you set them up.

 

 

Random boolean (specified) cutter:

[fae_alba]

 

 

I may start afresh and try to get the model looking a bit more like Humphrey/Papa Bear.

Alternately (additionally?), perhaps you can share a current version of Papa Bear and I can explore that. I think I have an early version of Papa Bear but you've surely changed him since that.

 

 

Rodney,

 

sent you an email with a model attached. Have fun with it!

[Rodney]

 

Random duckbox.

 

Can you believe this is my first time ever modeling a duckbox?

 

 

 

Testing SVG import (mostly).

[Rodney]

I was playing with transparent patches and light and noticed how nice the colors where transferred into the shadows and found myself wondering, "Is this something new in v19?".

Launching v18 and running through the same project I observed the same results. No, nothing new here.

 

So, why is it that I've not been using this 'feature' before?

 

 

Aside: One of the reasons I looked forward to the 'Duplicate' option arriving in v19 was to assist with lighting.

The idea being to set up a light in an ideal location and quickly be able to duplicate that light in the same location but with other settings. For example, one light might effect the surface of objects in the scene while the duplicate only casts shadows. Another example, having the duplicated light(s) just slightly offset from the original. Other examples abound.