Art Pipeline Proposal

[rivalin]

Some ideas, maybe someone in charge will listen.


Art Pipeline



Official Positions


Art Director (eg JackS) -provides the design briefs.


Technical Liason (eg Chuck Starchaser) -tells artists what to do with regard to technical stuff.


Art Administrator (eg Pyramid) -knows what models go where, organises the wiki, prioritises development.


Concept Artists (a senior concept artist) -come up with concept art.


Modellers (a senior modeller e.g. Fendorin) -model the ships and other assets.


Texture Artists (a senior texture artist) texture the ships and stations and so on.


Vegastrike Contributors (new developers) fill in missing simple textures and placeholders, do file format conversions.


Front End Guy (eg me) Does overlays, wallpapers, screensavers, renders, runs the ModDB page, a deviantart page, conceptart.org thread Thread, Game-Artist.net thread.





Here’s an example workflow;




Art Director;


design brief; human light fighter, solid mechanical design, heavily armed, quad winged design.

>>>>>>>

Art Administrator;

prioritises the design as high priority, updates the wiki accordingly.

>>>>>>>

Concept Artist;

creates some concept art based on the design brief




>>>>>>>

Modeller;

models the unit based on the concept art



>>>>>>>

Texturer;

textures the model, using guidance from the concept art.




>>>>>>>

Vegastrike Contributor;

takes the model and the texture, scales it, checks the alignment, converts it to BFXM and creates a units.CSV

>>>>>>>

Front End Guy;

creates the overlays during development, does renders of the finished model, does post processing, posts the pictures to websites, makes a wallpaper featuring it etc etc.





>>>>>>>

Art Director;

gives the model final approval and it goes into the game.


Anyone who blunders in with random ideas is simply directed to a sticky detailing what they need to do.

Keeping the pipeline linear and modular (i.e. every model needs every stage of the pipeline completed,no models without concept art first etc) ensures that certain problems that have plagued the game are dealt with; people can drop in and out of development at will, rather than a model being useless when the original creator gets bored half way (eg I got all the way to file conversion with a ship I was working on, got stuck there, and now in all likelihood it won't get into the game and was a complete waste of time), it also ensures that there’s concept art for every ship when they need to be remodelled. Having a design brief for every ship means that ships being reimagined with new concept art are understood properly (eg someone remodelling a ship without there being an original design brief may mistake missile launchers for fuel tanks and so on).




Anyway, I think I’ve made a decent substantive proposal that would be simple to implement and would give the art side of the game stronger focus and direction, it would allow people to come into the project knowing they could stick to their specialisation and not have to worry about doing everything themselves, it would be quick and easy to assign projects and keep track of where development is at any one time and it would best utilize scarce resources without allowing partial contributions to slip through the cracks. This system would create somewhere “the buck stops” so to speak; thus you would know that someone knowledgeable in the specific field could give you a decent answer to any question.


-I know what the rebuttals are going to be, they’re the same ones we see trotted out again and again, about how “organisation scares people off”, “ let’s just do everything informally” “that's already how it works, kinda” (it isn’t) . I don’t really think the aforementioned arguments hold water; the sort of people you want are not the ones that throw tantrums the first model they’ve ever made isn’t accepted into the game and storm off, the sort of people you want are pro’s using a little of their spare time, university students doing courses in relevant fields and people who work on mods for commercial games; namely people that are used to structure and organisation and specialisation. Many professional modellers won’t have come up with a concept themselves for years, many texture artists won’t have modelled for years, the point is that most of the people who have the ability to provide decent quality work don’t like having to be the entire pipeline themselves (doing that entire example pipeline was not fun) and they don’t like seeing art direction without strong leadership. Right now we have JackS saying yes/no at the end of the pipeline, we need leadership and direction from the beginning of the pipeline. Please consider this seriously, rather than giving the normal non committal response, then having the art component of the game continue working the same half cocked way it is now, that is, going nowhere, slowly. I’ve made an effort to contribute in a small way to the game before, hopefully that will qualify me to at least have what I’m saying listened to if not agreed with, and hopefully the fact that you don’t write code doesn’t disqualify you from having an opinion.

-cue the silent treatment/stonewalling

[loki1950]

Well thought out rivalin and with the more artist friendly tolls that have been proposed make a nice one two punch

Enjoy the Choice

[pyramid]

Excellent ideas and outline of the organization. I strongly agree with rivalin that organization rather simplifies things by making them modular and transparent in terms of information.

A pipeline would be a different thing, stating the process and what tools and approaches artists need to take into consideration at the different stages of the process.

Starting from this proposal, I would see the following steps to be undertaken for the start:
* Add the organizational structure to the project page, maybe under an own heading "Artistic Organization". I am fine with the proposal as it stands, though the technical liaison are potentially all active developers (e.g. chuck for shaders and textures, klauss & safemode for mesher and units.csv)
* reorganize the 3d models page to have columns representing the different stages of the process (guideline, concept, model mesh, textures, frontend, integration)
* finalize the texture requirements and pipeline (chuck) for the new shaders
* make that sticky post for new artists
* start filling empty places

[jackS]

Organization is something we direly need more of. That we don't have it doesn't mean we don't want it

I second Pyramids motions re: what we need to start doing in order to support the development of this process. I do, however, think the last two items (sticky post and start filling in) may need to be partially reversed in order such that there are at least some filled in examples before we tell all of the artists that "this is how things are going to work"

[pyramid]

The project page has been enhanced with the new chapter "Artistic Organization". Please review and post proposals for improvement.

Currently, I am working on reorganizing the 3d models page. While the previous "Vessels" chapter has been replaced with a new one containing a new table structure with more detailed and atomized information, the old chapter has been temporarily renamed to "Vessels (Old Organization)". Now, gradually, the contents of the old tables will be moved to the new table while I will confirm the status of each of the points for existing and being-worked-on models. When the transition is done, the old chapter will be removed altogether.
The new table already filled with some units. As it grows it will be consecutively divided into different sections to simplify the editing and maintenance.

[rivalin]

Good stuff I'll try to lend a hand with the unit lists, but my wiki markup's a bit rusty.

As for the organisation, I went back over it and made a slight adjustment here and there to future proof it a bit. I also did a little diagram to illustrate it that might look nice on the wiki page.




I also thought it might be a good idea to put a pm link next to each name so that people have one location from which they can contact whoever they need for help.

Sound acceptable?

[pyramid]

Since I've looked already a bit into the status of various vessel models, I'll be glad to continue on this task without interference.

Can't see the org chart from work, due to "company policy"@firewall blocking the venue. If you store the chart in the gallery (there is a special album for wiki documents), i can link it then on the page.

Wouldn't know how to make a link to pm from the wiki. Feel free to investigate an option and make it real.

EDIT: And as a bonus, those models that have "ASSIGNED" status, clearly state who has the ball atm.

[Deus Siddis]

Roster Page:

Cool, didn't expect to see my callsign there already.
You can put my real name, Jonathan Kelly, up there too. Sometimes I change my handle over the years, but not my actual name.

What does the texture artist distinction mean though? Fendorin does his own textures, I do my own textures, wouldn't we both be in this category or both not?


Models Page:

What is a dielectric map?

Will there be support in the future for dynamic localized damage textures, that show scorch marks and holes where in game weapons fire has actually come in contact with the ship? IMO, this would certainly look alot better than static damage maps.

If so, would this require an unwrapped collision mesh whose UVs rougly correspond to the the graphical meshes?

[rivalin]

The organisation is based on an ideal future setup based on the way most games are designed. Generally people are better at one thing than another, for example I'd love it if I could just model ships and someone would texture them for me etc etc. The ship I modelled isn't in game because I had trouble converting it (same as your model), with lots of specialised artists it makes sure that work doesn't get wasted.

This is all moot at the moment because we don't have any dedicated texture artists or concept artists but it's something to work towards. Basically people can do as much as they want, but when there are more artists in the future this setup will give people the freedom to stick to what they're good at rather than being forced to labour through parts of the artistic process they dislike.

for info about dielectrics check out Chuck's thread about his cinemut shader; I think it's something to do with the look of certain materials like metallic car paints.

[chuck_starchaser]

[OT:]

Dielectric constant is a property of materials. For practical purposes it's a property of non-metals. (The dielectric constant of metals is not given or even mentioned. Some physicist was saying that we'd have to use negative numbers in reference to metals' dielectric constants.)

The dielectric constant of a material affects its dielectric reflectivity (as opposed to metals, which have metallic reflectivity). Reflections off glass, or on the surface of water are good examples of dielectric reflectivity.

The difference between these two kinds of reflectivity is profound:

Metallic reflectivity has the color of the material. Reflections on gold look yellowish, for example. Also, the amount of reflection does not vary with the view angle.

Dielectric reflectivity is always "white", meaning full-spectrum. IOW, a colorful magazine's reflection on water is unchanged in terms of coloricity (hue or saturation). But the amount of reflection does vary with the view angle. Looking straight through a shop's window you see through pretty well, but if you put your head against the glass and turn it 90 degrees, at shallow angles the glass appears to approach a perfect mirror.

In this graph:



It's kind of turned around a funny way, but anyways, it graphs dielectric reflectivity as a function of dielectric constant and view angle. All the way to the right, dielectric k starts at like 1.1 (1.0 is problematic; produces undefined numbers), and ends at 10.0 at the left. View angle goes from normal to the surface at the far end of the graph, to very superficial and shallow at the nearer end. Thus, at the nearer end reflectivity is total, 1.0, regardless of dielectric constant. At the far end, 90 degrees, reflectivity simplifies to ((1-k)/(1+k))^2.

As I'm sure you can appreciate, we cannot represent realistic materials with a simplified model of having separate colors for diffuse and specular.
That's why models in so many games look like they are made of plastic.
The simplified shading models they use, as the old OpenGl model did, was only aproximately true for plastics. Paints never really looked like paint.

Besides, having separate diffuse and specular colors lends itself to having incorrect and impossible materials.

The CineMut shader represents materials using real physical parameters. There's only one "color" for a material; call it "diffuse" if you like; but you can vary the % of it that is diffuse and the % of it that goes to specular using a single texture channel. That's for metallic reflectivity.
For non-metals, such as plastics and glossy paints and glass, you simply specify the dielectric constant via another texture channel. You won't need to worry about how much it should reflect; the shader will compute the exact reflectivity given the view angle and dielectric constant, using the Fresnel equation, per-pixel.
Finally, some materials, such as molded plastics, have a blend of diffuse color and dielectric specularity; so there's another channel called "dielectric blend".

So, the 6 channels just described characterize a material as to its "chemistry". There's another channel, the shininess, which characterizes the finish, from highly polished/glossy, to very rough. Rough specularity is different from diffuse reflectivity.
And there's another texture channel to control amount and type of detail texture blend, to make speckled or wavey surfaces.

As for the question of localized damage, that's been explored many times and it's a herculean task to program.

The way damage will be handled in CineMut is via a special ambient occlusion baking with floating occluders, like:



You bake an ambient occlusion like that:



Which on the model would look like,



But, then, the LaGrande texture preprocessor and packer will take that and divide it by the plain ambient occlusion, to get something like,



From which it will compute normalmap modifiers, --"damage dU/dV" channels that are also part of the texture package, and that will be blended with the standard normalmap normals at runtime, in the shader, in proportion to damage.

The good news is you won't have to look up dielectric constants of materials in chemistry books; there will be a library of materials you can pick from and apply to your model in Blender, including many metals, plastics, ceramics, high, medium and low gloss paints and whatnot; then you bake textures, and then LaGrande will take those input textures and spit out textures for packing with the mesh file. Somewhere along that line, you'll be able to take the intermediate textures and add your art to them, of course. The bakings are only a way to get the basic materials right.

[/OT]

[Deus Siddis]

Dielectric reflectivity is always "white", meaning full-spectrum. IOW, a colorful magazine's reflection on water is unchanged in terms of coloricity (hue or saturation). But the amount of reflection does vary with the view angle. Looking straight through a shop's window you see through pretty well, but if you put your head against the glass and turn it 90 degrees, at shallow angles the glass appears to approach a perfect mirror.

Wow, I didn't know this could be done in a gpu in realtime yet.

As I'm sure you can appreciate, we cannot represent realistic materials with a simplified model of having separate colors for diffuse and specular.
That's why models in so many games look like they are made of plastic.
The simplified shading models they use, as the old OpenGl model did, was only aproximately true for plastics. Paints never really looked like paint.

True, but I kind of feel that way about alot of material types beyond even glass, plastic and paint. Organic stuff for instance, like animal and plant surfaces where the light is actually penetrating partially or entirely (example, sunlight going through banana leaves in Crysis trailers).

Besides, having separate diffuse and specular colors lends itself to having incorrect and impossible materials.

Well, in some mods, intelligent use of 'incorrect' materials might mean more 'alien' looking materials that enhance the game's visuals. So maybe fresnel and material influenced specular color wouldn't be a terrible combination for a shield effect or the surface of an alien hull.

Might be a good idea to let the artist choose these things if they don't look absolutely buggy and terrible, instead of modern realworld physics as dangerous as that sounds. Though haven't tried anything like this yet myself.

The CineMut shader represents materials using real physical parameters. There's only one "color" for a material; call it "diffuse" if you like; but you can vary the % of it that is diffuse and the % of it that goes to specular using a single texture channel. That's for metallic reflectivity.
For non-metals, such as plastics and glossy paints and glass, you simply specify the dielectric constant via another texture channel. You won't need to worry about how much it should reflect; the shader will compute the exact reflectivity given the view angle and dielectric constant, using the Fresnel equation, per-pixel.
Finally, some materials, such as molded plastics, have a blend of diffuse color and dielectric specularity; so there's another channel called "dielectric blend".

So basically, diffuse and specular will get replaced by the albedo texture right?

As for the question of localized damage, that's been explored many times and it's a herculean task to program.

Really? Because its been implemented in alot of games, including open source games, albeit to the environment instead of the actors (like ships or characters). And none of those games have anything even close to what you are talking about here.

From which it will compute normalmap modifiers, --"damage dU/dV" channels that are also part of the texture package, and that will be blended with the standard normalmap normals at runtime, in the shader, in proportion to damage.

Wow, so then it will look like layers of your armor have been ripped off clean. I wonder if you could make the edges where there is alot of contrast render in the glow shader as bright orange for a while, so that they looked close to molten.

The good news is you won't have to look up dielectric constants of materials in chemistry books; there will be a library of materials you can pick from and apply to your model in Blender, including many metals, plastics, ceramics, high, medium and low gloss paints and whatnot; then you bake textures, and then LaGrande will take those input textures and spit out textures for packing with the mesh file. Somewhere along that line, you'll be able to take the intermediate textures and add your art to them, of course. The bakings are only a way to get the basic materials right.

Sounds like a smooth pipeline. When would you guess this will be available? It will be open source right?

Finally, this is unrelated to global illumination and its two 'PRT-N' / 'PRT-P' maps mentioned on the wiki in the shader how to? I'm trying to get an idea for what the final list of maps for a 3D model will look like, if the GI is indeed a separate thing it will look like this:

Normal
Albedo
Glow
Shininess
Damage
PRTN
PRTP

True?

[chuck_starchaser]

Answered at the CineMut thread.

[Oblivion]

Rivalin this is indeed very helpful to churning out artwork. Heh. It speeds up the work a lot.

[loki1950]

Nice to see your avatar Oblivion

Enjoy the Choice

[Oblivion]

LOL, just visiting loki, Wanting to see the 0.5 screens!

[loki1950]

Have you had a look at chuck's shader dev thread looking very good

Enjoy the Choice

[Oblivion]

Where?

Anyway, my head's aching LOL, and I'm sleepy. I'll check tomorrow, I may model some of Fendorin's concepts if he doesn't mind. No promises! LOL

Nyt loki

[loki1950]

Here http://vegastrike.sourceforge.net/forum ... hp?t=11609 and
Nyt Oblivion

Enjoy the Choice