Direction of the Tool

Announcements and discussion regarding any projects related to Cyan Worlds' Plasma Engine including (but not limited to) Engine, Drizzle, OfflineKI, PyPRP, and libHSPlasma.

Direction of the Tool

Postby andylegate » Wed Mar 05, 2008 6:00 am

Wanted to open this discussion as I'm curious about several things, and this is also an opertunity to thank people too.

First I'd like to really thank all those that have been involved in any aspect of Tool Development. Both for the plugin and for the ULM. You guys have done and keep doing a wonderful job. It's a lot of hard work, that you don't get paid for, and I don't think you get quite enough thanks for what you have done, and the gift you have given aspiring Age Creators. So on the behalf of many I'd like to give y'all a big standing ovation!

Now, I'd like to ask those of you that are hot and heavy in the Tool Dev, exactly what do you see for the future of the plugin? Meaning how it works with Blender of course.

Here is why I'm asking:

I've noticed myself, and have heard other, "seasoned" Writers mention this too. Which is the ability to get Blender to render what we see in Uru, also in Blender. It seems to many of us that we seem to end up having to do more and more "guess work" when it comes to painting and lighting up our Ages. That the only way to find out how things look, is to export and load your Age up.

Let me elaborate upon that further. Again, I'm NOT belittling the work of those who have dedicated so much time an knowledge on the plugin, heck no! You guys have made Age Creation more and more powerful over the last few months than I think anyone expected. Again, my hat is off to you.

However, let me go into what I'm seeing here:

How we paint our world:

Currently, the way we use Blender and the plugin tool, we have no real time picture of what our Age is going to look like until we export it and link in. About the only sure fire method of "What you see in Blender is what you get in Uru" is if we simply give our meshes colored materials and leave it at that.
Well, unless you are taking some very cheap drugs, this is not how the real world looks! Everything has textures of some sort, not flat colors. This method is really only good for helping get an idea of the geometric shape of things to come in your Age.

We can't use Blender's texture lab unfortunately. Or actually I should say, we can't use it to apply some of the great textures it can make to our meshed directly and have the export into Plasma.
You can use it to make your own textures however, by setting up a flat plane, giving it the materials and textures you want, and then having Blender render it into a picture you can save, and then use that same picture as your texture for UV Mapping.
Kind of the long way around, but no different from using a seperate texturing too (Filter Forge), or paint shop program to make your own stuff. It takes time to cruise the web to find textures you can use, or take your camera outside and snap off photos that you might need. So that doesn't bother me too much.

However, when we go to apply these textures, "What you see in Blender is NOT what you get in Uru."
You can just straight apply your image with UV Mapping, and be done with it. And you'll end up with an Age that looks like Camp Bravo Day Time. It's obvious in there what is wrong, because it doesn't convey the realism we constantly strive for. Most things in the real world don't simply end sharply and the next thing start sharply. They tend to blend on into another. The exception to that is man made objects of course, but when it comes to nature, just about everything blends into each other.

Right now the way we blend things together in Blender, depending on how you do it, you either end up with something that will sort of look like it's suppose to in Uru, or you end up with something that doesn't look like it at all!
Okay, well, we can just hit the Render button in Blender to get an idea, right? Wrong.

This is the point of contention: Lighting. There is such an extreme difference in how lighting works in Blender as opposed to how it works for the plugin. You have to mess around with your lighting quite a bit until you get what you THINK it should look like in Uru.
But before you export, you'd better re-adjust that lighting, or when you do and link in, you'll be in for a horrible surprise: nothing looks like it's suppose to.

Instead we are having to resort to guess work again. Export the Age, link in, try to see what may or may not look right. Quit. Go back into Blender, try something different, re-export again. Link in again, look again, ah crud, still not right.......
When it would be so much easier to simply tell Blender to render that scene and look right then.

Yesterday, my wife downloaded the latest and last pod cast from TCT. During their Round Table discussion, in which the pretty much griped about everything, but one thing was brought up, the lack of realism in how Uru looks compared to older Myst games like Riven.

Well, we know one reason for that. Riven wasn't a 3D enviorment that you walked around in. It was 3D all right, and was rendered into many scenes that were then presented as high resolution (for the time) graphics picture to the player.
Well good news! That was 10 years ago. We've come a LONG way dudes! Technology has done nothing but get better and better.

It seems to me, that in many cases, our Fan Ages (User Ages, Explorer Ages, whatever y'all want to call them) some things look much better than they do in Uru itself.
Theoretically, it would seem that you could create textures and lighting so good, that it will convey a sense of realism that we see in other Games.

But not if we have to do guess work all the time.

Imagine what would be possible, if while working on your Age, you could go in and set up your lighting just right in Blender and render it there, knowing that how it looks there is how it will look in the Game. I know you can do it now, but it takes a great amount of fiddling with the sliders. Ask Nadnerb and he'll tell you.
Imagine what would be possible, if we could have textures look the way they do in Uru also in Blender before you export the Age. Imagine how you could do that finely detailed work right there, instead of having to guess and hope that it comes out right.

So here it is. Blender is open source. What is it to keep Blender from being modified itself? I've seen this disussion before on here. One of the excuises not to try and do this is due to the fact that newer versions of Blender are always coming out.
Okay, but unless it was a completely new version that does things completely different, I don't see the problem. If it changes that much, I imagine our plugin wouldn't work anymore either. It too would have to be rewritten.

So I guess what I'm asking is: Is this possible? I know that even if it were, it is something that would take a VERY long time to do, partly because of the size of the job, but also because everyone is doing this in their spare time. It's not something that we can demand, but only request and see where it goes.

While I'm a complete stranger to the technical aspect of this, and it may not even be possible, it is still something that I wanted to bring up.
Keep in mind, the easier things are to do in order to make Ages, the more people you'll have actually creating Ages.
How many new people did we have flock over here upon the announcement of the closure of MOUL and then the videos we did to promote Age creation and testing? A lot.
Now, how many of those have given Age Creation a try, only to get bogged down in the confusion and guess work it takes for texture blending and lighting?
We don't even have a tutorial for lighting. And for good reason. It's just about all guess work. I mean you can follow Trylon's First Real Age tutorial, but after that, when it comes to your own creation, you have to come up with your own lighting, and that isn't easy without some guidance or instruction.
How many have given up do to the frustration of this? We might not ever know.

I still think y'all have done a damn good job. I mean everything we have has been researched, reversed engineered, and made by all of you, at no charge to us! Can't argue with that! Y'all have done a incredible job, considering everything.

Many other game companies put out mod tools for their games because they know how it extends the life of them. I know Cyan wanted too, but for whatever their reasons were not able to (I think it was going to cost way too much back then, and of course now it's legal entanglements). So I'm really proud of the fact that you all were able to come this far.

So in any case, where do we go from here? Is what I'm talking about even possible? Is it way too complex for anyone to touch?
Or is it everyone is happy enough as it is and Andy should shut up about this? :D
"I'm still trying to find the plKey for Crud!"
Blender Age Creation Tutorials
3DS Max Age Creation Tutorials
User avatar
Posts: 2348
Joined: Mon Oct 01, 2007 7:47 am

Re: Direction of the Tool

Postby Aloys » Wed Mar 05, 2008 7:04 am

Being a tool user I off course agree. :) (especially seeing how I started that Uru Blender discussion some weeks ago)
However. As you mention, the guys are already doing a great job with all this. Spending so much time and energy in it. Much more than we could ask for. So I simply dare not ask for more. :)
I agree that the lighting and texturing aspect has become a real problem since version 1, but I don't really see what could be done. Trying to do anything like the Uru Blender idea might be a solution in that regard, but that would mean re-writing part of its renderer and that's a lot of work. Is it worth it? I don't know. Depends how much time would need to be spent doing it. I think doing an Uru Blender would be a great advance as a whole, and not only in the mapping/lighting department. It would be much more user friendly, we would be more productive with it etc.. We've already been through this discussion. Again: is it worth it? That's a difficult question. Personally I think it would be; but the decision would be in the programmers' hands..

Finally what I think something that could be done in a more immediate fashion would be to expand the tutorials to better explain how to deal with the current features and their shortcomings. (But as I said previously; since I can't spend the time myself to write any of those tutorials, I'd rather avoid any backseat driving. :) )
User avatar
Posts: 1919
Joined: Sun Oct 21, 2007 7:57 pm
Location: France (GMT +1)

Re: Direction of the Tool

Postby Chacal » Wed Mar 05, 2008 11:58 am

Couldn't this be done more simply in a future version of the exporter?
Start with Blender's normal lighting and texturing tools and parameters and translate them into Plasma parameters during export?

"The weak can never forgive. Forgiveness is an attribute of the strong."
-- Mahatma Gandhi
User avatar
Posts: 2463
Joined: Tue Nov 06, 2007 2:45 pm
Location: Quebec, Canada

Re: Direction of the Tool

Postby Marcello » Wed Mar 05, 2008 12:39 pm

Stuff like realism are partly a matter of using the right texture, amazing lighting and in most games nowadays cool DirectX shaders. I personally love some of the screenshots I've seen with blownout hightights and stuff. Cyan doesn't use DirectX to it's fullest because they wanted to be the game as downward compatible as possible. If they would make a realtime 3D game now with an engine like that of Unreal, Crysis, Doom or whatever... I'm sure they can come pretty close to Riven like images.
User avatar
Posts: 374
Joined: Sun Nov 04, 2007 8:59 am
Location: Haarlem, The Netherlands

Re: Direction of the Tool

Postby Nadnerb » Wed Mar 05, 2008 2:12 pm

Chacal wrote:Couldn't this be done more simply in a future version of the exporter?
Start with Blender's normal lighting and texturing tools and parameters and translate them into Plasma parameters during export?
There have been attempts in the past, and there will be more in the future, to line up blender's lighting model with plasma's, and more accurately translate between the two. Unfortunately, testing this sort of thing requires a large amount of back and forth with exports, and whenever changes get implemented it destroys everyone else's already finely tuned lighting. :P

Also, while there are some things that can be done to bring the plasma output closer to what you see in blender, there are still some particular difficulties in getting the lighting in blender to "match up" with the lighting in plasma. For one thing, it is possible to set up blender's renderer so that it's output looks almost like what you would see in plasma, but doing so at the moment requires having the user manually set up a large number of default values on all of their materials. Most people like to leave things at the defaults if they don't have a specific effect in mind, and most don't know what all the right settings are, much less want to set them for every material and object they create. We can't change blender's default values without actually making our own version, and we're trying to stay out of that for now. Blender upgrades really do have good stuff in them, and it would be a shame to leave that behind just so that we can use different defaults. For instance, some of the newer texture baking options (which could be used in lightmapping, for example) were only added a few versions ago. What's more, the blender team is actually moving toward increasing the ease of use of blender itself by exposing all of the functionality that was originally only available through hotkeys (Read the history of blender. It's a lot easier to learn now than it was before, and that's saying something. ;)) in menus and other user-friendly interface type things.

Also, the current direction of the plugin does in fact have ease of use in mind. From what I can tell, a good amount of work was put in making the stencil material output in plasma as close as possible to the internal behavior of blender. This involved swapping the layers around, (they're the other way around in plasma) and calculating alpha values in the same way blender does based on greyscale images. (plasma reads alpha values for stencils, not greyscale) So you can rest assured that we are trying to do things "the blender way" as much as possible. Nobody is intentionally making this thing hard to use.

I'm not quite sure how to wrap this up, but I'd just like to say that you can rest assured that we do consider the 'ease of use' factor when implementing features and upgrading old ones. We're not looking for recognition, just trying to make age building possible as possible and hoping people don't complain too loudly when we give them new stuff. ;)
Live KI: 34914 MOULa KI: 23247 Gehn KI: 11588 Available Ages: TunnelDemo3, BoxAge, Odema
Posts: 1057
Joined: Fri Sep 28, 2007 8:01 pm
Location: US (Eastern Time)

Re: Direction of the Tool

Postby Trylon » Wed Mar 05, 2008 2:41 pm

andylegate wrote:Currently, the way we use Blender and the plugin tool, we have no real time picture of what our Age is going to look like until we export it and link in.

Well, they don't call it "the Art" for nothing eh... ;)
One day I ran through the cleft for the fiftieth time, and found that uru held no peace for me anymore.
User avatar
Posts: 1446
Joined: Fri Sep 28, 2007 11:08 pm
Location: Gone from Uru

Re: Direction of the Tool

Postby Kato » Wed Mar 05, 2008 3:21 pm

Correct me if I'm wrong, but, with the exception of lighting, isn't the "Textured" view in Blender (not a render) very much like the final output? Obviously, vertpainted alpha doesn't work and all that--but all your textures are there, your lighting is there, your animations are there...I don't see any problem in building an Age based on the "Textured" view in Blender.

(explorer card designed and created by me)
User avatar
Posts: 315
Joined: Fri Sep 28, 2007 8:02 pm
Location: South USA

Re: Direction of the Tool

Postby Nadnerb » Wed Mar 05, 2008 3:46 pm

Unless your materials have more than one layer, and you are using stencils and addition. Then you have to render.

I used to avoid multilayered materials and do everything in textured view with decals. Worked quite well. Then the new version of the plugin came out and multilayered materials became so useful that I just couldn't stay away.
Live KI: 34914 MOULa KI: 23247 Gehn KI: 11588 Available Ages: TunnelDemo3, BoxAge, Odema
Posts: 1057
Joined: Fri Sep 28, 2007 8:01 pm
Location: US (Eastern Time)

Return to Plasma Development

Who is online

Users browsing this forum: No registered users and 1 guest