Is there a head of the texturing departement yet? i think people would be much more willing to commit if there was a todo list like in the modeling dpt and maybe some style guides for the textures. i would love to do some texturing work, but i'm usually not very organized.

So:

1. i think someone should be appointed the manager in this forum
2. i'd love to see a style guide by the art director (William Sutton?)
3. the manager should compile a todo list
4. more people need to sign up!

I agree Marcel. For whatever reason, no one will volunteer for this position. I thought finding animators would be difficult, but finding a texturer has been impossible. Isn't it strange that we have 1000s of frames of Woot without bump maps, eye lashes, or skin color? You would think some talented artist in the A:M community would see our glaring need and hear our appeals for help and step right up to be "Texture General." However, we are patient people, and we wait optimistically for lightning to strike.

Are you guys talking about decaling materials... or both?

One of the things that has caused me to avoid this issue is having to deal with the actual flattening and placement of the decals. It is not one of my strengths... despite what you see from my stuff... I tend to be lazy and just stamp stuff all over the place.

If someone else can do all that... hard stuff... create a template for the texture needed... maybe with a "pre-decaled" model for testing this might be more encouraging for those who can do textures... but hate all that other stuff that goes along with it... or are not as proficient at it... I can't flatten worth crap to be honest.

Just a thougt.

-Vern

You're a hard-nosed negotiator, Vern. But... Okay, you can be "Texture General." (Where do I send the funny hat and ribbons?)

Alright, who will "flatten" for Vern? Come on, guys... You don't have to be Michael Angelo to flatten. You flatten - Vern paints, and wa-la, we have decaled models, (sounds like Vern can apply materials without your help - apparently that part of his brain still works).

I really think simple skin is fine for this sort of project style (and magnitude). I don't think we need to see the pours or anything like that. Maybe a bit of cheek blush or wrinkles in the older characters.
And are lashes really necessary? Many great characters have none.

I'm gonna disagree Ken. I think the characters deserve much more than we've given them so far (I'm still hopefull V14 will include SSS so we can use that). But, like Marcel pointed out, we really need a style guide to go on. The original direction was "like Shrek" which would mean a lot more texturing and a lot more detail but it's still pretty vague. At the very least their faces need more details. I'm more than willing to help out with texturing, I don't even mind flattening or painting. But I've been holding off jumping in here and some of the other areas like lighting for the same reasons Marcel listed.

Kevin

I understand what you're saying but I don't completely agree. I agree that we probably don't need to go to the extreme of painting pores (although if someone was to do that I would certainly not mind) but we need some painting of the skin color with the subtle differences of hue on the cheeks, chin, nose and forehead at least. That would already make a big difference from the current uniform plastic/vinyl look. SSS alone would not cut it. It would only change the vinyl look for a wax look.

We also need a minimum of wrinkles around the eyes in particular and some bumps for the lips as well as some specularity maps for those areas. We also need wrinkles and color maps for the hands.

The clothes need some texturing too. In scenes where we have close-up shots of the characters, clothes surface texturing is important otherwise they seem to be wearing rubber suits. So at the minimum, we would need some bump textures for rendering the fabrics surfacing for close-ups.

Again, for close-up shots, lashes are necessary IMO. Lashes play a good part in eye expressiveness especially with blinks and when the eyes are shut. When the eyes are shut, the absence of lashes make the eyes look like meaty slits not too nice to look at. The lashes mask that. And once we have lashes for close-up shots, we might as well use them for the other shots too.

I think there is already enough talent in this thread alone that we could come up with some style guides. I'd suggest to anyone interested in getting involved with texturing, to start doing a research on different styles. Not just using previous animated films as examples but doing a broader search including painters and illustrators for examples and submit that for discussion on a TWO texturing thread. The research and documentation phase is what any AD would first do anyway. Lets get something started in this department.

I say don't hold off. Anybody.

Let's take the animation as an example. AFAIK, nobody really wrote an "animation style guide". Yet, we see new scenes being animated every day. And they are getting better and better while the animators are acquiring experience. We should be able to do the same thing with texturing.

Let's get texturing started. Take one character. Work on some texturing. Post a render of the result. Get feedback. Refine. Etc. Let's just start texturing and get better with experience. The TWO project is, in big part, about developing talents. The first textures will very probably be redone as experience grows anyway.

I'm currently working on lighting Act 1, Sequence 2. I would like to see Woot being better textured. That would be a good start. By working on already animated and lit scenes, the texturing task should have clearer goals and purposes which should help decision taking.

Fair enough Yves (some good points), since I'm already going to be working on hair for Woot and Tin Servant this weekend I'll take a stab at doing some additional texturing on one (or even both) to see what I can come up with.

Kevin

3DPainter isn't far oh-offf......

I just thought that until the rig and animation was done it might be advisable not to texture yet. You never know if more splines need to be added to, or taking away from, characters until then. Won't you?

Again, my suggestion is that we start developing texturing expertise and knowhow right away. Develop the techniques, the skills and the craft. Test and refine different texturing ideas and styles. Basically, develop the texturing expertise at the same rate that we develop the animation expertise so that when animations are completed we already have the texturing expertise.

I see no harm at doing texturing on current models. Even if models eventually change, we will only need to reapply the textures. Reapplying the textures is still much quicker that painting the textures and getting them right given some rendering context. Let's not wait untill we really need the textures before we start developing the skills.

I think the decals should be very realistic. It is easy to put a ton of effort into a high res decal so they look really really real. It is amazing how much style and realism is possible with a simple model and fantastically realistic decals...

Each character could have a palette of colors specific to them i would think as a start. This should be determined by the art director or whatever. Just an idea. Then if people are creating textures they stay within a set style.

Maybe some scenes have different lighting and would require unique textures specific to the lighting for that scene?

One of the things that also concerned me was...

... all the different ways to apply decals... everyone has their own way of doing it...

Flatten pose?
Flatten action?
No flatten... just stamp till your fingers hurt?
Manual flattening? Partial flattening?
Cylindrical decal on "round" characters? (I've done that... tricky but it can work).?
etc...

If we have a bunch of people creating the decals and each person is flattening or decaling a model on their own... you will end up with 50 different styles of decal application...

That is why I thought it would be best if a defined "style" of decal application be "locked down". Each character has a "wireframe" decal template. This could be done right away.

------------

Not to draw any comparisons... please ignore any other subtle inuendos or negative comparisons to this following statement...

... In a "cookie cut people program" that starts with a "P" and sounds like "Loser"...

There are standard image maps that can be changed in Photoshop. They are all the "same size" and "layout". You don't even have to understand 3D at all to make new ones... just open the template and paint.

------------

Another example... on the Extra CD... there was Jim's romance novel cover art... er... that Queen of the Leopards project.

All of the decaling was done. If you wanted to change the look of one of his characters you could just paint over his decals. You didn't need to flatten anything to get started. No stamping of decals... no preliminary work... just paint a clown face on the queen and laugh and laugh... uh... sorry.

I find often that the flattening and decaling process is WAY harder than actually creating the darn decals. I get 90% done... apply the decal... and realize it is all wrong and I need to flatten it differently... rearrange splines... I hate that part.

Creating the textures is fun. I enjoy it.

With these characters... yikes... If I have to flatten them myself... forget it.

Some of you seem to be able to do it in your sleep. It is a real chore for me.

I would think at this point some sort of "library" of "blank" decals could be started for each character. If they change later... the wireframe "template" can be updated and the textures could easily be applied.

I would be better off... and get more done... if I could just work in Photoshop... maybe do a test render on the character in the default pose (with a proper choreography lighting set up... I suck at that too).

I am rambling...

-vern

Good points. I think flatening in an action is the way to go. Doesn't make the models bigger and you can come back to it if need be.

Also, what's the best way to flatten a head? Is the nose separate? The ears should be.....the back of the head is easy enough. Then there's going to be a need for textures to define the hair areas. They'll probably be separate textures altogether.

Hands are quite easy too....maybe some hand textures can be used for several characters....probably the ones only briefly seen.

Does that flatten feature actually work? How?

All these things need ironing out.

Well, for the latest character I am working on I decided to put alot of detail into the facial textures. I didn't worry about the back of the head, but I included the ears in the flatten. Here it the template I made for that character. I also included a small version of the final color map that I used with my template.

The flaten feature still works. But being an automatic function, it is not perfect. Still, it is a good start. Further tweaking can be added to the CP after the initial flatening. That is what I do.

One important point, when flatening a mesh, is to keep the relative area of the patches as much as they are in the model as possible. Of course, patches on the periphery of a flatened mesh will be relatively larger than than patches in the center of the flatened mesh. But the important point is to keep nearby patches relative areas as similar as the unflatened mesh as possible. In other words, if one patch is twice as large as the neighbor patch, don't attempt to make them about the same area when flatened. This will produce texture details discontinuity all over the model.

Also, there was a mention of 3D painter. This tool can work with already flatened decals or produce its own UV maps. I don't advise using the auto UV maps feature. This will not work when MipMaped. No texturing tool that does that sort of auto UV mapping (this includes Z-brush) will produce nice results with MipMapping. Even with 3D-painter, it is still better to produce a flatened map first.

I agree with your concerns Vern. I think we need to develop a decaling technique that will be applied to all our characters. This is especially important for characters where decaling faces, heads, hands, etc can already be tricky. Since more than one person will probably have to work on the decals, it would be better to have one uniform way of doing that so when one takes a characters and wants to modify the textures, everything places are already well known and expected.

This said, to get there, we really need a decaling champion. One that will take on the task of developing the technique and documenting it. Just like the skeleton and facial rigs were developed by champions who took on the task and developed superb tools for everybody taking into account feedback from the users. We need to get into that kind of development process for decals too.

I don't know how hard this will be and since I am not familar with other productions of this nature... but...

I would think that creating decal "templates" should be like... the last thing done. I would almost say that no more tweaking of the models should be done after that point.

There should be a "drop dead no more tweaks it's done" moment... then decal templates could be created by who ever is responsible for flattening the models.

Just a thought... wouldn't want to hinder perfection by not allowing changes to the model meshes... but changes could go on forever... there should be a time to stop model tweaking.

--------

I really love that UV editor tutorial... the "famous" one with the ballrog thingy?... who did that? Was that Will Sutton? Or someone else? I forget.

Anyway... I loved the layout of the head decal. That kind of thing for each character. The whole character with one or maybe two or three big decals to paint from a template with the wireframe and maybe even indicators for parts of the body... whatever.

Maybe one for the head... one for the body... depending on the character. I guess clothes would be a different thing... but similar.

This allows you to "see" quickly how everything "looks" together while painting. It keeps everything simple in "one spot". Less file management issues.

... uh... just to be clear, I don't practice what I preach... I am not usually that effiecient with this type of thing.

----------

If you decide to use the "3d painter" gizmo... not sure about that... I think it is Windows only, correct? If so I that would be an issue for me... uh... I probably wouldn't be inclined to use it as much. Plus it is going to cost money right? Maybe Hash could buy a group license for the worker bees. Just a thought.

-vern

Yes, that is Will Sutton's talent and genius on display.
He's still got the tutorial on his website:

http://www.zandoria.com/uv.htm

Warning: Artistic nudity ahead (Not the Balrog though... he's PG13!).

For planar decals I would suggest a flattening in a pose would be better than flattening in an action. That way the flattening stays with the model (it shouldn't inflate the model size by very much at all), and if the mesh is modified then it will be easy to position any new CPs/splines in respect of the decal (i.e. in UV-space).So *that's* what you were doing during your hiatus from the forum!?Surely you have a copy of Jim Talbot's excellent tutorial on your hard drive? There's a link to it on the ARM. Watch it a few times, then try it out. It truly is an excellent tutorial, and you'll end up wondering why you found it all so difficult before!Yes, but you do it so well.

I see the point. It is not the first time this is brought up. But don't quite agree.

A face decal "template" for a given character should stay the same no matter the changes, like adding more splines or suppressing some. It should stay the same because the general morphology of the face, the head or the body should stay the same too. If splines are added it should only be a matter of revisiting the flatening action(*) and adjust the newly added splines and possibly adjusting the flattened position or adjacent splines. This should be fairly easy to do especially if the first decal "template" was saved from a screen grab.

IMO, if a character is modified by adding more splines or suppressing some, then revisiting the flatten actions to reposition the added splines and possibly their surrounding splines and reapplying the decals would anyway represent far less work than revisiting the smartskins, fan bone weights, animations related poses/actions and choreography actions. So if the "Let's wait until modeling is finished" argument was applied to smartskinning, posing and animating as well, we would be stuck in a deadlock.

With this, I agree. There should be a stop point. This said, my experience with several other projects is that the stop point, for some outstanding situations, can extend as far as the final delivery date. Sad but unfortunately true. Nevertheless, we must fight the tendency to seek perfection. This is an absolute necessity. Perfectionism can quickly suck up the available resources (time, human and money) of a project. So for any task, there should be a "perfect enough to stop" point beyond which, any further modification must be surveyed and approved.

(*) I prefer actions for flatening because it does not increase the model file size. Actions files used for flatening should be saved in the character folder though. Not in the general Action folder IMO.

Unless a modification is deemed critical, perhaps it should be simply noted. Sometime towards the end of the project they can be revisited. This way all the accrued modification requests can be done at once. This would minimise disturbance to all the people using those models during production.

Please download an updated woot as I've altered his face mesh in preparation for adusting his face poses. There shouldn't be need for any more changes after this.

Just checking if someone have already started flatening and or texturing Woot because if nobody did it, I thought I might at least flaten Woot's face and supply the action for someone else to start texturing it.

So anybody started flatenig Woot face? Kevin?

Or anybody interested in doing that? Jeff?

I created a flattened action of his face this weekend (and added hair). I haven't really created much in the way of maps for it yet but I'll try and get things updated tonight. Where do you suggest saving the action?

Kevin

Cool!

Since a flatten action can only be used for the Woot character I suggest saving it along with the Woot model file.

Well crap, does anyone know of a way to copy a model with hair styling intact into another model? Due to my screwup with SVN I endedup with a Woot model with just the skullcap/hair, which I attempted to copy into the real Woot model. Unfortunately when I tried I lose all the styling (about 3-4 hours of work). I could redo it quicker now that I've figured out a technique but I'd rather not if I can avoid it. Anyone have any ideas? Thanks.

Kevin

Dang.....maybe it's possible to drag the grooming info into the new model? Not sure.

I think you just import that skullcap/hair model into the other model... you know... open the model with the old hair... rt click import... new hair.

I guess you should delete everything else not needed like bones and stuff. I think this works...

-Vern

Do it the reverse - import the woot model into the skullcap/hair model

tried that, the hair isn't included at all, so I tried opening in the both models in the same project and copying and pasting into the Woot model. This time I got hair but all the styling was lost...




I could try that but I'd be scared I'd lose something in the existing Woot model. Anyone know of any issues with doing this? Guess it couldn't hurt to try...

Nancy's solution should work.

This has come up before and that was the solution... I got it backwards. Hair styling can not be copied and pasted.

It would seem to me that if you import the model into the hair model... all of the bones, rigging and relationships should go with it....

if it doesn't... uh... that would be a 6 legged organism with a hard exoskeleton... there's one crawling on me as we speak... eewww.... gross.

-vern

I actually had this same problem a while ago. I imported the model where I had the hair and it worked well.

Alright, cool, I'll try it tonight, hopefully I won't screw up Woot too bad...

You can not import the woot model into the hair model. The reason for this is that the percentage pose sliders will default to 0% to 100%, most of them are custom percentages. This is why when installing the squetch rig you need to import your mesh into the rig and not the other way around. Sorry.

Kevin, you Masters of Computer Science, you: you should be able to figure out the XML file format, do a file compare in your editor, and do a "by hand" copy and paste.

You had to go and bring that up Martin, didn't you

I'll try the "manual merge" process and if that fails I'll just redo the styling.

Kevin - I wish you great success. I've many times wanted to copy hair grooming data (and have done it with rigged/posed critters the convoluted way of importing into the hair model - tho I have not tried in ver 13, with squetch rig).

It would be really appreciated after you've done this you might enlighten & share with us your findings.

I am curious as to which items one would be manually editing? the pose percentages in the xml? - or is it being implied that the hair grooming data can be manually copied & pasted into a new model ?

Thanks

Well, my experiments with editing the xml were not succesful (sorry Nancy). I just couldn't figure out what needed to be brought over to properly include the grooming. I tried doing a diff on the two files but that wasn't very helpful. I tried importing the hair guide settings under the material after importing the skull cap but it didn't seem to work either since the cp's had changed(?). I could probably spend some more time and figure it out but I've decided to just redo the grooming, at this point it will probably take me less time. I'll try to get at least the hair and the flatten action into SVN tonight, after messing around with the xml I ran out of time last night.

Kevin

What happened to cause this? Maybe it might help avoid it in future.

Again out of curiousity - Did you try importing Woot into the hair model & then changing the percentage values of the poses in the xml to reflect what they were in the Woot only model? Wasn't clear from your explanation (nor do I know if this is all that would be needed, aside from boning the hair patches).

And yes, after spending all this time to save time (been there, done that, still doing that, will do it again) - is probably better to just bite the hair-of-the-dog bullet.

Just my own stupidity.



No, I didn't try that, might be worth a shot. I can always diff the files to hopefully see what pose values got reset.

Here's what I have done in the past...first, in a text editor, copy all of the pose slider information in the file (everything from the end of the "</MESH>" to the beginning of "<BONES>") and save it to another text file so you don't lose it (I usually just keep it open). Next, import Woot into your hair model and save the model. Then, open the new Woot model in a text editor and replace all of the information between "</MESH>" and "<BONES>" with the text file you saved from the original. That will give you back all of the percentages that were initially on the sliders and everything should work fine.

This has worked for me on several occasions, I haven't posted it anywhere before because I wanted to discourage anyone that didn't know what they were doing from trying it....I really don't want to get those "you messed up my model" messages.

Good luck.

Well, that didn't work either. I was able to get Woot and the hair back into the same model but half the groups were missing and none of the materials were moved over. I'm just too worried about screwing up someone elses work. I've decided to just redo the styling...thanks for all the help anyway

What happens if you create a blank choreography, put both Woot and the hair into it, and then export the choreography as a model?

Back to the texturing process discussion:

*I'd like to see flattening done in an action to keep the data out of the model file. The action files can be saved in the Actor's folder in an actions subdirectory. i.e. Woot's flattening action could go in Woot/Actions
*We should plan on having to re-apply textures at some point down the road due to unforseen circumstance. Flattening actions will be a part of that. I think we should also consider adding groups to the actors to hide geometry for decaling purposes.

I'll get to work on a todo list for texturing. I'd love to see a style guide too but I think the style is going to develop as the work is done (like the rest of the movie).

As for the process, how about this: Flatting the model will include creating a flattening action, defining a group to hide appropriate geometry and stamping a blank image to create the UV map (would the artist want to do this last step himself?).

Vern, Marcel and Kevin: Keep talking here and with me so we can work out a process for this. If I know what you need (flattening actions and so on) I'll make sure you have it so you can do your work.

Rhett

great!
should we create a small tutorial outlining the steps of the technique along with instructions where to put what and naming conventions?

should the whole model be flattened into one action and therefore be stamped with exactly one decal (like its typically done for games) or should there be separate decals for different body regions? former method would allow re-applying the texture with one step, while latter method will allow for mixing and matching different textures (like dirty clothes, scratched arms, etc)?

I prefer breaking up the models into different regions to stamp, it's easier to make higher res maps. I've started doing exactly what Rhett mentioned to Woot. So far I've just flattened his face in action and created a group for the flattened cp's. I'm not sure it matters who does the stamping (though I kind of prefer to do my own). BTW, how much latitude do we have in doing this texturing? Meaning, can I just go to town on texturing Woot? Also, is there some reason we are using 20% ambience on Woot (and I assume the other characters)?

Kevin

No No ambiance. This messes up all effort for good lighting. So if you find ambiance anywhere in any model, you may set it to 0% right away.

That's what I assumed, just wanted to make sure before I turned it off. I was suprised to see ambiance set all over the place on Woot. In fact I probably spent the first hour tweaking my hair material because it kept turning out really bright and I couldn't figure out why.