Game Engines!

[GGTharos]

Sure there are things left to be improved, but there's no need to look at other engines (or more to the point, they're not useful for DCS).

 

Have you read these forums at all? Haven't you read what's coming with Nevada?

[HiJack]

support for the 11 :D

[Trident]

1) CryEngine 3

 

Not suitable for a variety of reasons, such as draw distance and physics (it does rigid body interaction extremely well, but that's not the same thing as aerodynamics and flight mechanics). Other aspects (AI, sensor modelling, LOD handling) will similarly be optimised for infantry-simulation as well.

 

2) Outerra

 

Outerra is, as has been pointed out already, little more than a landscape rendering environment at the moment. Nevertheless, the fundamental approach used for generating additional mesh detail is interesting and could help solve some of the problems faced by ED in improving their current terrain mesh. However, implementing the Outerra engine as a whole would be prohibitively time consuming, since it is so basic in its current state and would require a huge amount of work to bring it up to speed.

 

3) Rise Of Flight Physics Simulation Engine

 

Rise of Flight is a very impressive flight sim engine, but you're giving DCS too little credit. It does almost everything mentioned in that video just as well or better (systems: electrical, hydraulic, fuel), the one area where RoF is indisputably ahead is structural damage modelling - in that regard it is definitely the industry benchmark. It also has very clever 3D trees and excellent flame effects that I would love to see in DCS, but other than these things I don't see what the fuss is all about - especially considering how much effort it would take to make the engine capable of handling modern aircraft.

 

Then 'environment' isn't going to be there until computing power increases some 10000-fold. Then maybe they can start doing something. Every time you double your view distance, you need at minimum 4x the computing power. Think of it as difference between low, medium and high scenes and tree vis radius as well.

 

You have /no/ idea what you're talking about. No idea whatsoever.

 

Credit where it's due though, the best tactical FPS engines are rapidly approaching the point where the drawdistance would be entirely suitable for a helo sim. That is pretty impressive!

 

Regarding that top link...Wtf? That looks photo realistic....Surely no mere mortal PC can run that?

 

There's not actually anything terribly special going on there. Texture resolution isn't really any better than what we have in DCS, they're just well-executed photo-textures (i.e. accurately geo-referenced, colour corrected and cleaned of artifacts - which is more than you can say about most other photo terrains). However, they still seem to contain shadowing from the relief and will therefore look slightly off if you change the time of day too far away from the hour when the source material was taken - changing lighting conditions is something DCS will handle much better. Nonetheless, the Switzerland Professional add-on is definitely among the best photo tile sets ever released, despite the fact that it's almost three years old now (almost no image artifacts such as clouds, distortion or objects that aren't represented in 3D by Autogen, stuff that ruins most other photographic texture add-ons).

 

The primary reason why the terrain looks so amazing in that video is that the add-on also includes a new 3D model for the Swiss landscape, with a resolution that is *vastly* superior to DCS. That's basically all there is to it, apart from the photo textures being presented very flatteringly by using the correct lighting conditions.

 

Having said that, M$FS does have one very nice feature (apart from the high resolution mesh, which DCS could probably handle too) that isn't really visible in this video: dynamic terrain selfshadowing. In pretty much every other way, the DCS engine is already light years ahead though.

 

Not in any practical way. Guess why.

 

Considering that 95% of the eye candy in that video comes from the better terrain mesh and bearing in mind the polycounts DCS already handles thanks to its extremely detailed aircraft/vehicles and dense buildings/vegetation, I'm not actually so sure about that.

 

In short, you're dreaming if you think it is possible to make DCS have a 60 km view radius and render every leaf as in Crysis while having the great systems, battle and weapons modelling of current DCS.

 

Correct. To add one more reason to what you already mentioned, it's also a question of sheer manpower. The more detail you have in each square kilometre of terrain, the longer it will take to create said square kilometre with equal development resources. Now imagine how hard it would be to build a theatre which encompasses many tenthousands of square kilometres at a detail level which matches a Crysis map with perhaps 30 square kilometres (assuming for a moment that performance would not become an issue).

 

There are ways to somewhat mitigate the impact for relatively modest improvements in the level of detail, but going for FPS-style detail is definitely out of the question for manpower reasons alone, not to mention other problems.

[GGTharos]

Please tell me which apps use your rig facility fully. :)

 

Easier said than done, though DCS keeps splitting threads off, so at least it uses more than just one core. :)

[GGTharos]

Ok, obviously, I can safely say this only happens in your dreams, but it isn't a bad dream ;)

[jmod]

Ok, obviously, I can safely say this only happens in your dreams, but it isn't a bad dream

 

We have to remind that, lots of human inventions were started by dreams!

And when the dream comes true, it's not bad licensing TFCSE, something like Prepar3D.

[Cali]

We have to remind that, lots of human inventions were started by dreams!

And when the dream comes true, it's not bad licensing TFCSE, something like Prepar3D.

 

That is true, everything starts from a dream or a vision. Maybe in a few years someones dream will come true and they'll find a way to fix or create a new game engine that will fix a lot of the problems we have now. Look at how much games and computers have improved over the last 10 years.

[Boberro]

Have you read these forums at all? Haven't you read what's coming with Nevada?

 

What's coming with Nevada which could be called "super" by user? New textures with same old resolution which are blurred (and often empty not filled by buildings)? or maybe angular, like set square mountains? :D

 

it is good move but ... show me pros of Dx11 in DCS for the end user, not for programmer.

[Pyroflash]

Tessellation, enhanced DOF, a new API that allows for certain functions to optimize code. There are many more advantages of DX11 to the end user of course, this isn't a pushover update like DX10 was.

[Succellus]

I think we could buy F-35 Glass instead!

 

LOL!

[OGREMAN]

The NEW ULTIMATE SIM ENVIRONMENT

 

Hello Gents,

For all those flight simmers who get frustrated by the "pc plastic" world graphics that is common in todays generation of flight sims, I would like to draw your attention to what I belief is the next big thing in not only flight sim, but all serious simulations. Procedural graphics engines are where its at for the future flight sims.

With procedural graphics you can say goodbye to the world of POP UP everything.... no more ... entire landscapes poping into existance or objects changing their appearance with dramatic suddeness. With procedural graphic generation like that of Outerra, if an object is large enough to be represented (ie one pixel) it is always visible and the only thing that would fade out its detail is naturally applied atmospheric light scattering, mist or other weather conditions.

 

 

 

 

 

And if that is not enough to get flight simmers thumping the table for change, how about these tasty trinkets of information...

• The graphics engine in Outerra is GPU driven, leaving the CPU free to manage the modeling of the SIMULATION DETAILS and not the terrain.
  
• The terrrain represented and avilable encompases the whole of planet earth and its detailed down to as little as 20cm resolution. (no more limited land areas to cramp your game, you can choose to fly anywhere).
  
• The graphic detail and quality of the visuals is without comparrison in any simulator. The true texture and detail of rock & mountain surfaces are faithfully reproduce in true 3D instead of those wallpaper images draped, warped & streched over obvious geometric meshes.
  

The Outerra terrain generating engine is now within a very few months of being available to Developers and casual users, check it out here http://outerra.com/

 

"There must be a down side" I hear you say, well yes there is a problem. The code and even the basic structure of todays flight sims is not directly compatible with a proceedural terrain engine like Outerra.

Existing flight sims would have to be re-written in new code and that means a lot of work.

But wait a minute! when FSX,MSFS,DCS and all the others put out their "latest offering" its often not backward compatible anyway and they are constantly re arranging the terrain graphics in an attempt to improve that aspect of their simulation. So why would they not take the plunge and commit to leaving the present generation of flight sims in their wake?

It seems to me that the "down side" is that a developer would have to do some rewrite work ..... GO check out the link and its graphics, then I will be interested to hear if you think that such a commitment would be worth it.

[sobek]

GO check out the link and its graphics, then I will be interested to hear if you think that such a commitment would be worth it.

 

I don't need to, because i know by the time a complete rewrite is done, most independent studios would have thrown in the towel for a lack of income. So, no, it's not worth it.

[OGREMAN]

I don't need to, because i know by the time a complete rewrite is done, most independent studios would have thrown in the towel for a lack of income. So, no, it's not worth it.

 

Nice Sobek, thanks for that .... very open minded and enlightened view. I do hope that level of consideration is not the norm in flight sim developer circles .... or we are all doomed to "fly in aspic" never see a truely natural looking flight environment.

Edited June 24, 2011 by OGREMAN

[sobek]

That's not me being pessimistic, besides, i'm not a dev. Sure i'd like to see lots of sims pop up with new engines. But i'd rather see what little developers we have left not go out of business than risk getting no sims at all for the prospect of using a new engine.

 

Look at 1c, they just came out with a complete rewrite. If they didn't have ubisoft to back them up financially, they'd probably face harsh times with their release of CoD, maybe even bankruptcy.

 

Besides, as the current situation displays, sim devs are uneager to use engines other than their in-house designs. To accuse them of ignorance, lack of farsight or entrepreneurial spirit is going a bit far, don't you think?

Edited June 24, 2011 by sobek

[OGREMAN]

Re Read the comment .... it was an expression of hope, not an accusation.

[Ali Fish]

regards procedural driven. i absolutely agree.

 

@ Sobek re 1C, Is 1C seperate from the 1C:maddox company, did 1C:maddox have financial trouble. isnt that the route cause for everything problematic over on that side ?

[GGTharos]

Because it would mean throwing away their current product essentially, meaning wasted time, effort, and money.

 

Further, Outerra simply isn't ready, and it isn't like GPU-based computing is your greatest friend ever, either - don't think for a moment you won't suffer slow-downs or other issues. The GPUs are barely keeping up in some cases as is!

 

And once more: The terrain engines in various sims are intricately integrated into the game itself, including AI pathfinding algos, collision algos, who knows what else - not to mention the terrain modeling for roads, etc. It is a /massive/ undrtaking to 'take the plunge'. To say that it's 'just a bit of a rewrite' is simply ignorant (and I mean that in the nicest, most literal way of the word).

 

So, as far as I can (make an educated) guess, where DCS and Outerra are concerned? Not happening. Period (at least, not within the next 5 years if ever). ED is busily upgrading its own terrain engine.

 

But wait a minute! when FSX,MSFS,DCS and all the others put out their "latest offering" its often not backward compatible anyway and they are constantly re arranging the terrain graphics in an attempt to improve that aspect of their simulation. So why would they not take the plunge and commit to leaving the present generation of flight sims in their wake?

It seems to me that the "down side" is that a developer would have to do some rewrite work ..... GO check out the link and its graphics, then I will be interested to hear if you think that such a commitment would be worth it.

[Conure]

Looks beautiful. I've no idea about programming etc, but I sincerely wish our sims looked like that :)

[OGREMAN]

My Simulator Heaven! Outerra terrain and DCS Aircraft & game modeling.... drool. IMHO the worst aspect of flight sims today is that we all turn a blind eye to those ridiculus terrain graphics, its like we have simply gone into denial about how truly bad/un-natural they look, and here we all are seemingly afraid to admit that we want better.

The argument that producing a new sim is somehow throwing away all that previous development is silly, I can still buy Flanker,Flaming Clffs,LOMAC, MSF and use them, they did not dissapear just because a new sim became available.

Edited June 25, 2011 by OGREMAN
Add a comment

[Trident]

If ED can take advantage of terrain tessellation with their upgrade to DirectX11, procedural terrain like in Outerra might become redundant, as far as I can see. If they wish to continue using a conventional terrain engine however, procedural mesh refinement could be used to speed up improvement of the existing mesh rather than editing the elevation model manually, as mentioned before. I. e. rather than implementing this approach in realtime, they could take the current mesh, generate an improved version with a fractal algorithm, fix any artifacts by hand and then render it with the current engine. While that throws away the benefit of dynamic LOD computation (which enables Outerra to display sub-metre mesh-resolutions), it would probably be the quickest way to match the best M$FS add-ons.

Edited June 25, 2011 by Trident

[jmod]

So, as far as I can (make an educated) guess, where DCS and Outerra are concerned? Not happening. Period (at least, not within the next 5 years if ever). ED is busily upgrading its own terrain engine.

 

I think they're just busy reinventing the wheel!

Look what's going on FSX:

Edited August 5, 2011 by jmod

[sobek]

I think they're just busy reinventing the wheel!

Look what's gong on FSX:

 

So? A few missiles and explosion effects bolted on to FSX do not a combat simulator make.

[GGTharos]

Who's busy reinventing the wheel? The guys trying to turn FSX into a combat sim? I agree.

[Trident]

I think they're just busy reinventing the wheel!

 

How so? It would take far more effort to bring the FSX engine up to scratch (since ED's engine is already superior in pretty much every way other than terrain mesh detail) than to improve DCS to the point where it matches or exceeds FSX in the landscape elevation model department. In its current state, Microsoft's engine is wholly inadequate for a DCS-style combat sim and would require a huge amount of work to prevent it from being a step back in most respects (sensor modelling, flight modelling, systems/damage modelling, AI, multiplayer and yes, even most aspects of the graphics engine).

 

While I agree that ED's terrain mesh resolution is in dire need of an upgrade, using the FSX engine (or, for reasons already discussed, any of the other ones suggested here) is certainly NOT the answer. Although VRS is planning to do more than bolt a few explosion effects onto the basic sim (saying that definitely sells them short), it will probably take them years to produce anything comparable to DCS, for all the same reasons mentioned above. DCS is the product of evolutionary development over a period spanning about 15 years, by a very talented team!

Edited July 1, 2011 by Trident

[Smokin Hole]

Outera. WOW! Time for Sims to Leave the Stone Age

 

This from the X-Plane 10 blog. Unfortunately it's not part of X-Plane either.