DirectX 11

As lately nearly everyone seems to use Windows 7 and seems to have Hardware supporting DirectX 11 features, I would actually really love to see gamestudio to take the step to an additional renderer using D3D 11.

For most users it should not really change anything and those that want, could go crazy with stuff like geometry shaders and tesselation and just do some really advanced techdemos or whatever comes out of such experiments.
And at least at the moment, it could still be possible to get it done before Unity or Torque do.

Last time I read such a post here, I laught and thought what an idiot is posting such bullshit. But to me it now seems to be a good time to bring it up again, especially as you are moving away from DirectPlay and DirectAudio (or whatever it is called), which should finally make this switch not THAT much work.


And some question:
Is there a way to assign a material created at runtime to level geometry? If not, this is another feature request .

Thanks

The biggest benefit would be the promotion you can do with DX11 support.

I think that for conitec, mobile platforms like android have higher priority at the moment (which is also great!).

It's very unlikely to have an Android port (jcl has mentioned that serveral times before), because it's no DirectX based plattform....

Where did you get that?
Quotes by jcl:

While I of course wouldn´t mind gamestudio to support mobile devices, I would currently just use Unity for this (at least if I hadn´t my own engine :P) and I somehow doubt that gamestudio will be able to seriously compete there...

Engine is slowly getting rid of platform dependent libraries though. First DirectPlay and now the sound engine. The next main thing is the graphics renderer, mobile platforms would require an opengl es renderer.

Does Acknex even support DX10? Am asking because I don't miss any specific DX11 feature (though tesselation and bokeh would be cool), but some of the DX10 capacities. But as those come hand in hand with DX11 compatibility, I support Slins request.

dx10 is pretty much unfinished version of dx11.

like vista being unfinished version of 7

@Quad: If i remember back..how long they used DX9 compared to DX10 o.O. Had a "bad" feeling about DX10 when they suddenly announced DX11.

not to mention, despite the existence of dx11 and dx10, dx9 is still the main version big players use. That is kind of about the capabilities of the current consoles though. They can't use dx11 featrues there, so they dont bother wasting resources on using these features on pc. We will see dx11 as main dx version when consoles get dx11 features.

According to Sony, the PS3 is the generation for the next decade, so we still have to wait five years.
Might I request a DirectX 13 renderer then?

read aum100.

Instead of pushing dx11 it would be better if they would start to face the opengl direction. It's true that opengl is not equally powerful to directx and it's another thing but it's also true that opengl opens up a bigger chance to target multiple plattforms. It's a big work to port an directx engine to opengl especially if you don't want to restart coding the whole stuff. but at some point it's better to take this mammut work to gain more reachable users later. Immagine Linux Users if they had a ready to use toolset to create games. There are a lot of opensocure engines, libraries etc. but the targeting group of people which Gamestudio has, is the oposite of those typical linux users. Not everyone likes to make 100 Steps to finally have ready a working engine. Gamestudio would have a big boost with opengl in numers of users. ( with linux users I mean the ones who just have a easy to use system like ubuntu, I have different ones on my machine and I know the booring side of linux and game engines for linux) plus opengl would open the world to mac too, and with webgl es ( which is a difference but not that much compared to dx) even browser and mobile plattforms...by the way shader coding is not that different between directx and opengl. The enduser would not even note that big difference if they use lite-c. the only one who would note it, would be JCL when porting the base engine code to opengl with the differences in deeper details when it comes down to 3d things and their calculation and rendering. but let's be honest, would not opengl open more doors to Gamestudio? I love Windows but it's not said that we all will use it in some years like we did today and yesterday..even windows can lose attraction for customers and developers. Just a matter of time and what the competitors of the operating system will do in future. better to start right now to work on it, then to do it on the last days...the fanbase of gamestudio would probably still use it, but others would think twice then. hope to read in future anytime that acknex will go the opengl direction.

Yes. DirectX 11 could be quickly implemented, but is not very important at this time. DX 9 is still more important. And we will need OpenGL ES earlier than DX 11.

And personnaly DX11 compared to DX9 or DX10 is not a big big change like DX7 compared to DX9 for example !
I also don't feel the urge to DX11 , or perhaps some people targetting AAA Games usings last technology ?
Other things are more important, like also the Iphone possiblity
mentionned in the AUM !

Just my opinion !

while i can see both sides, dx11 is pretty cool and yet android support would be amazing (for me at least) dx9 to dx11 is a hugh jump. with dynamic tessalation, we can sculpt objects, bake the maps, make a low poly version, and have almost exactly the same look as the sculpt in realtime, including full silhouette. imagine having lod that doesnt just pick the detail in texture and some premade models to having full procedural lod.

Actually DX11 is more than only tesselation. It can speed rendering up a lot especially in combination with deferred rendering. Many tasks can be put into one single step. This allows almost unlimited lights in a scene and many postprocessings at the same time. There is no doubt that this is the future and that deferred rendering will beat forward rendering on DX11 systems and on the future consoles generation too.

i would prefer multiplatform support instead of dx11
especially with the power of lite-c gamestudio would gain a bunch of new customers i think

Exactly my opinion.

this is all true, but speed upgrades arent a big deal to me. they are expected in an update. bleeding edge features dont always make the cut though and im glad in dx11 they did and apparently work properly as well unlike dx10

What is that "power" you see behin Lite-C? I fail to see this being an all selling feature. Sure it's not crap but let's do the comparison to e.g. Unity. There you have Javascript and C#. Both more or less are general and popular languages (although as I understand it Unity's Javascript has minor differences) with great editors available and of course lots of good documentation and tutorials. This also covers one sort of easy language and one more advanced language. In terms of difficulty / complexity I don't see Lite-C being too far away from something like C#...

That's why I don't see the benefit (and Unity isn't the only engine out there offering an easy script language and a more complex C language be it C# or C++). I actually tend to see Lite-C as a disadvantage for 3DGS. In my opinion it would have been better to introduce an easy scripting language rather than creating Lite-C and leave the option for e.g. C++ for the pros. Especially as 3DGS tends to be an engine for newbies and lone wolves this probably would have been the better choice in the long run. Well sure there were things like that Atari cooperation but I have to say I never heard anything at all concerning Lite-C and Atari besides the initial news about the partnership...

Concerning DX11 for 3DGS I actually tend to agree with jcl. It's nice to have but currently of no great importance (even AAA usually titles don't feature it recently and don't count in stuff like AVP 3 or Crysis 2 where the DX11 support is barely existant and consists pretty much of "We use tesselation here and there")...

"In my opinion it would have been better to introduce an easy scripting language rather than creating Lite-C and leave the option for e.g. C++ for the pros. "


This would create a too big hole. I, myself, would never ever ever use a language that would be much simpler than liteC, also C++ is too hard on the other side.

LiteC is great!

(also,i dont care for dx11/android)

Same here, personally I don't even care for console support. Imagine A8/ A9 supported consoles, then you still had to become an official XBox/PS3/Wii developer and I highly doubt that many games would make it there. We all know how difficult it is to complete a single PC game. Android (/iPhone) would have been a good platform for indie developers ~2 years ago, but today it's just overcrowded.

My wishes for the (near) future: New WED, new homepage, new art assets, new (additional) Gamestudio developer (maybe even rehire Marco Grubert, I cannot remember why he has left).

I think that's more important than DX11.

actually superku, i never played your game but i watched some videos, and your game would probably be a great xbox live/tablet/smartphone game. if wii becomes indie friendly ever, than using the accelerometer in the wiimote/phone/tablet seems like it would be a fun involvement. again never played though, just from what the gameplay looked like.

More batches per frame, direct access to the depthbuffer, extended HLSL and lots of other cool things are in my opinion worth a month of work. It however should not replace DirectX 9 but should be usable alternatively, which could also lead to a clean abstraction needed for OpenGL ES 2.0, which then also shouldn´t be too much work, except the shader system, but then one could maybe just assign two effects to a material and the correct one is then used depending on the renderer.
Btw, to the shader haters around here: Did you know that you can´t render anything in OpenGL ES 2.0 without shaders?
The really big problem there however is probably Lite-C... It would have to compile to another or even several other architectures. Or it needs to be parsed and executed at runtime, which however will most probably take quite some speed out of it. So maybe no Lite-C on Android?

I just see more potential in DirectX11 than in Android support. A new WED however is anyways the most important at the moment and some good way to debug Lite-c...
Also, wasn´t there somewhen the request to implement PerfHUD support? Should be easily implemented (I am however not sure if it is a good idea to have it always usable.) und would help a lot if it comes to more complex rendering techniques.

Crysis2 had no DX11 until 2 days ago. Crytek delivered the DX11 update, with some really nice effects. For example local Realtime reflections.
[video:youtube]http://www.youtube.com/watch?v=osIj6qbXY5g[/video]


Greetings
Rackscha

What's that realtime local reflection thing about? That's nothing new (just a mirror, right?)

Which afaik is no DX11 relevant feature though. At least I read a posting which was quite convincing and concluded, that tesselation is the only real DX11 feature - all the other things (which unfortunately often make next to no visual difference to the DX9 shaders) all were possible with DX10...

That's why I made my point: Most AAA titles don't support DX11 in the first place and many of those which technically do just use a feature like tesselation so they can put the DX11 stickers on their titles...

The first game that really makes use of DX11 might be Battlefield 3 as they seem to really have integrated it in the engine rather than using it as an "update" or small "add-on" here and there...

As Frank already pointed out, DX11 is not used in current games as those still need to support DirectX 9 for Xbox360.
Also I don´t see it as a reason against DirectX11, that it isn´t yet widely used in current games...
But if jcl decides to focus on other things in the near future, I can´t change that anyway, and I also am not sure if I want to, as he most probably has good reasons for his decisions.

I didn't want to emphasize this as a reason "against" DX11. It's just that the DX11 integration offers nothing despite a few nice-to-have features and some possible performance improvements (which on the other hand would require adaptations of the renderer and shader code so integrating higher speed-ups is no fire and forget task that gets done if you just support what DX11 can do). So the only argument left would be an integration because it's a standard nowadays and everybody uses it / expects it...

That's what I wanted to express: There's no such need of a DX11 integration as hardly anyone uses it (properly). This wouldn't make DX11 support a bad thing but if there are better / more valuable things to do those should be preffered...

One video called it "SSR", which I imagine is screen-space-reflections. I think Little Big Planet does a similar trick when you're near glass surfaces. This way you can get decent looking reflections on arbitrarily shaped surfaces.

Jibb

Who in indie game making will really be able to produce a big AAA today games to really need an update for DX11 features ?

Why not instead try to stabilize or make a solid base of shaders with why not extensions (shade-C) directly plugged to work with future versions of A8 ?

I know some people kinda strayed off topic, so here are my 2 cents..

I think we should always be looking to improve the game engine, but it seems to me what is really lacking is the art pipeline.....look at MED...let's either get rid of it or actually upgrade it. I do everything in blender and then fight with or use some quirky workaround, to get my models into GS...and even then I do most of that via code.....so I am not even concerned with the current state of wed.

I was wondering what that was, to me it just looked like a typical rtt reflection added over it.

I would've thought so, too. I can't remember where I saw the screenshot that called it "SSR", but I think we're getting used to what "SS" means when we're talking about shaders, and it's not the first time I've seen such an effect.

It's all quite pretty, but as mentioned before, the tessellation is the only thing we can't do in DX9 (although I have no doubt many of the other effects are more efficient with DX11).

Jibb

DX11 against DX9 has to be viewed from the point of a platform strategy. If an engine has to be cross-platform, then it makes sense to stick to DX9 as long as the next consoles generation appears. But if you have a technology based on Windows only then by all means, DX11 is a big improvement. You never can have almost unlimited lights with the current forward renderer. And DX9 implementations of deferred rendering are often too slow.

DX11 not only gives you tesselation, it can give you freedom where you can have much faster rendering, more lights and post-processing at almost no costs. It can be an advantage on Windows systems.

Of course many current games do not come with such modern technology because pc sales are only 10-20 percent of all sales. Consoles are more important. But a PC-only engine could use this situation as chance to fill a gap.

That was basicly my idea, but that Deferred rendering is problematic with DirectX 9 ist something I don´t think, although in DirectX 11, it is of course an advantage that one can for example directly use the depthbuffer and don´t have to generate your own depthmap. Lightprepass rendering, which is what all those current deferred rendering implementations with those high light counts are about, is actually kinda THE technique for good results on the current generation of consoles.
But yes, DX11 would be a great step forward. Performance and feature wize. I think that there is more possible than we can imagine at the moment, as so far no game even slightly used the given possibilities.
And again, also don´t forget the power of geometry shaders. They can be great for things like fur, but also for example arbitary formed bookeh effects.

geometry shaders are an incredibly powerful thing also for other rendering techniques like point clouds or voxels.

Thanks, haven't read that before.