DirectX 11 class graphics cards can enhance meshes on the fly, on the GPU via tessellation.  We wanted to shift the DSF elevation mesh toward raster data so that we would have the full source raster to feed into the GPU.  In this configuration, we can make a low resolution mesh, give the graphics card the full data and say “go to town.”

If the graphics card can ‘enhance’ the mesh quality, this solves a problem we have now: there is no rendering setting for mesh complexity.  Right now everyone uses the same meshes, so we have to limit mesh detail to meet the specs of low-end supported computers.  With GPU-enhanced terrain, users with more powerful systems can crank up the detail.

We’re not ready to code this yet, but one first step was modifying the DSF format to be ready for tessellation.  We did this with the X-Plane 10.0 global scenery, and a nice side effect was smaller DSFs. :Party:  :yahoo:  :yahoo:

 

http://developer.x-plane.com/

Great News!! :Applause:

The idea is not so new   ... but good to bring it up to the attention from time to time

 

http://developer.x-plane.com/2011/08/dsf-gets-raster-data/

Not sure what it all means, but it sounds good. Anything that can put that GTX 680 card of mine to work is good news :-)

Dynamic Hardware Tessellation:

https://developer.nvidia.com/content/dynamic-hardware-tessellation-basics

Dynamic Hardware Tessellation:

https://developer.nvidia.com/content/dynamic-hardware-tessellation-basics

 

Will this improve the aesthetics of the terrain ? Do you know if this something they plan on coding for X-plane 10

Its like Ben says ... its planned ... in the sense of a "plan" (it was already a plan / idea, when Ben came up with raster DEM data in the DSF the first time ... the link I pointed to above) .... but its not worked on now as far as I(!) know.

 

Its also - kind of - stated in the comments of the latest blog entry:

http://developer.x-plane.com/2013/04/dsf-mesh-formats-v9-vs-v10/#comments

 

I do not know what the ‘cost’ of tessellated terrain will be in terms of fps – since we don’t even have a prototype right now, we have no initial data.

 

And yes, of course this would (and will) increase the landscape aesthetics.

I only know at a basic level what tessellation does and in demos it looks pretty nice. Its nice applied to objects in first person games, thats mostly what I have seen.

 

I see problems applying it to landscape though. In a game with fantasy landscape I dont see a problem but recreating landscapes that people are familiar with is a different story. How could it improve on what we have?

 

The tessellation will smooth the surface out. Is that what we want? Is that the right thing to do? The way tessellation works really well in games is with displacement maps. Whenever you have seen a really good demo using the technology it would have had a displacement map which xplane does not have.

 

 

Here is a quick link explaining how it works, not far down you can see pictures of it working with examples of basic tessellation and then with displacement maps.

 

http://www.nvidia.com/object/tessellation.html

 

 

Reading it again, perhaps this is where it could be of most benefit.. Not so much in re-contouring the environment?

 

 

Seamless Level of Detail

In games with large, open environments you have probably noticed distant objects often pop in and out of existence. This is due to the game engine switching between different levels of detail, or LOD, to keep the geometric workload in check. Up until this point, there has been no easy way to vary the level of detail continuously since it would require keeping many versions of the same model or environment. Dynamic tessellation solves this problem by varying the level of detail on the fly. For example, when a distant building first comes into view, it may be rendered with only ten triangles. As you move closer, its prominent features emerge and extra triangles are used to outline details such as its window and roof. When you finally reach the door, a thousand triangles are devoted to rendering the antique brass handle alone, where each groove is carved out meticulously with displacement mapping. With dynamic tessellation, object popping is eliminated, and game environments can scale to near limitless geometric detail.

Jason, you are missing an important little piece in the puzzle.

 

The current mesh (the triangle size / density) in X-Plane - especially the one in the Global Scenery - is really "coarse" (partly because of size limitations in a multi DVD distribution) compared to what even the "average" 90m resolution (1201x1201 pixel per 1x1 degree area) DEM could deliver in detail. Even though X-Plane even uses an irregular mesh (which means, put more, smaller triangles where there is more elevation detail , and less, where everything is flat(ish) anyways), the default triangle sizes are usually too big to reproduce all available info ... And with the new XP10 feature of having the full elevation information in raster form inside the DSFs always opens the door, to have a better reproduction of them ....  as soon as triangles are smaller (the mesh is denser). Just take my HD Mesh scenery series which - on average - put in 2-3-times more triangles .... and this scenery is still just barely able to capture the details of the 90m mesh! Now, this is exactly the place, where the tessellation would help us .... namely, dynamically increasing the mesh resolution (without the need of more triangles in the DSF files - saving file size!), have a way to fully (or at least almost fully) utilize the information in the DEM data, and have nicer terrain. There is even more potential, when we (or someone else) puts higher resolution DEM data (like 30m .... with 3601x3601 samples) inside the DEMs ... then the tessellation could really bring out a lot of fine, nice details.

 

So, the idea is not so much the classic - like in many games - way of tessellation, where you "make up" detail (by some fractal algorithms etc. etc.) but to really put detailed - real - information to good use in the landscape visualization!

Sorry but this sort of stuff is not what XP needs right now.

Why oh why are the developers talking about refining the mesh when there are:

 

1) no seasonal textures

2) view / distance issues especially at high altutude

3) weather / clouds system needs major upgrade

4) vast majority of major airports with no buildings

Etc etc ad nauseum

 

Please, let's talk about refining the mesh for XP11; XP10 mesh is fine, there are much more important things that need fixing first. X-Plane will never be major league if the developers keep trying to fix stuff that doesn't need fixing!

 

My 2c

Why oh why (sorry, I usually try to resist ... but sometimes ...): is this topic NOT about the other missing stuff, but plainly about a new, future, planned (its not even in active development) feature, which people discuss about?

Why must every second thread be hijacked? There are enough threads about you - valid - points ...

While tessellation is bordering on being old news in the wider world, its good to see this type of tech finally hitting Flightsims.

 

 

Why oh why (sorry, I usually try to resist ... but sometimes ...): is this topic NOT about the other missing stuff, but plainly about a new, future, planned (its not even in active development) feature, which people discuss about?

Why must every second thread be hijacked? There are enough threads about you - valid - points ...

Hi Andras,

My point (borne out of the frustrating snail's pace development of XP) is that we occasionally hear these techy tidbits from the XP developers, but no real news that they're working on something that really needs addressing. It honestly makes one wonder where LM's priorities are. Sorry no hijacking intended - I just want XP to be seen to be heading in the right direction.

Well, for this feature especially, the land mesh rendering and capabilities can be offloaded from the CPU to the GPU, thus creating more headroom on the CPU for calculating other features. They may be at 90-100% capability resources, but this would take them down to 70% or so giving them additional resources to be dedicated elsewhere and enhance the performance of the product. It doesn't pay to have a sim built of 20 50% complete features. One built of 10 100% developed features runs better and has less room for error than one that is just slapped together and issues handled as they hinder enjoyability.

Andras, if the sim has more data available to it that isnt being used and tessellation can use it then thats a good thing.

 

    I didnt know if the sim had more data which wasnt being used and my concern was that tessellation would be used to 'infer' or make up data points that might make known locations less accurate.