Hi.Displacement-mapping is one of the more prominent features of PS3.0The trick with displacement-maps is that you can spare thousandsof polygons by using displacement-mapping.Why?With displacement-meshes you alter a given mesh. In your example,the 9k triangels are just reconfigured to form the bumpy structureinsteadof the perfect sphere. The Problem is, it still consists of 9k trianglesthat have to be rendered and lit.Displacement-mapping makes the use of so many triangles obsolete.The basic sphere can be rendered using only a few hundred trianglesand still look round if properly shaded (e.g. Phong).Displacement-mapping now extrudes the texture-map with theinfo stored in the displacement-map to shape the more complexstructure.In theory, that saves lots of geometrie-work.
Practicaly themath for calculating displacement-maps cripples that advantage,IF it is done by software on the CPU. Thats why your exampleshows that disp.mapping takes even longer then disp.meshing.However, the new cards can do it several orders of magnitude fasterin hardware. That way, lots of displacementmaps can be renderedby the card in realtime, saving geometry-power for other objects.
Originally posted by MadMonkeyHi.Displacement-mapping is one of the more prominent features of PS3.0The trick with displacement-maps is that you can spare thousandsof polygons by using displacement-mapping.Why?With displacement-meshes you alter a given mesh. In your example,the 9k triangels are just reconfigured to form the bumpy structureinsteadof the perfect sphere. The Problem is, it still consists of 9k trianglesthat have to be rendered and lit.Displacement-mapping makes the use of so many triangles obsolete.The basic sphere can be rendered using only a few hundred trianglesand still look round if properly shaded (e.g. Phong).Displacement-mapping now extrudes the texture-map with theinfo stored in the displacement-map to shape the more complexstructure.In theory, that saves lots of geometrie-work. Practicaly themath for calculating displacement-maps cripples that advantage,IF it is done by software on the CPU. Thats why your exampleshows that disp.mapping takes even longer then disp.meshing.However, the new cards can do it several orders of magnitude fasterin hardware.
That way, lots of displacementmaps can be renderedby the card in realtime, saving geometry-power for other objects. Where to begin.OKDisplacement mapping VS. Displacement mesh's:As i understood displacement mesh technique, it is somehowcomparable to vertex shaders. This realy modifies the 3D-meshof an object, creating a new one. It's like a template that can beapplied to any mesh, processes it and thus creates a new one.Once Processed, the new mesh can be viewed and rendered without additional workload.Displacementmapping however doesn't touch the 3D-mesh at all!It recalculates the extruded texture new for each frame. This costsenormous calculatingpower, but saves memory and bandwith due to much less complex 3D-meshes. So if a card has the feature andthe power to do displacementmapping, it is preferable to usingdisplacement meshes.
NV & ATI use displacement mapping.Displacement-mapping and shadermodels:Displacement-mapping is possible with PS 2.0 already, but therecent cards didn't pack the punch to use it regularly in games.The performancehit limited it to special applications and techdemos.GF 6800 and Rx800 will supposedly handle it much better.Additionaly, PS 3.0 feature an improoved implementation ofit and speed it up a bit.NV & ATI & support for shadermodels:Both, ATI & NV have features in the current cards (NV3x & R3xx) that exeed PS 2.0 specifications.NV however does not support all PS 2.0 instructions. The currentdrivers even for the GF 6800 line of cards still lack some PS2.0 shaders required by DX9.ATI supports all PS2.0 instructions required by DX9 on their recentand new line of cards!So, both can claim having PS 2.0 +, while ATI's claim is more justified.disp.mapping, disp.mesh & quality:In general, a disp.mesh should result in a better quality object.It creates a real, new mehs with the shape wanted. It will rendercorrect in any view and shouldn't produce any glitches.The downside is it requires ALOT more triangles then disp.mapping!Using disp.mapping can create odd effects, glitches, blurry edges,coarse reflections and distorted shadows on the 'surface', dependingon how accurate the pixelshader works. 16bit will likley procudesome visible errors, 24 bit will produce a mainly correct imageand 32bit will, if at all, only produce methematical errors not visibleand be almost undistinguishable from the disp.mesh while savingmemory due to way less triangles. You're so lost dude, Displacement mapping don’t do ANYTHING to the polygons, it’s just an effect making it LOOK like there are more polygons, and therefore you can make a brick wall or a rocky surface with only 2 polygons (which are minimum for a flat surface). Basically it makes a flat surface look bumpy.It's kind of an evolution to bump mapping, making you see the bumps from the side and make the bumps self shadowing, you are comparing the new displacement mapping techniques with old ones in CG programs. And it’s just not the same.The whole point is to make more detailed scenes without too much performance impact.
It does make it easier to make a brick wall for example, but that’s not the main reason it is used.What you describe it as, is no more then height mapping, as used to make the terrains in for example UT2003/4. So it’s not the raw rendering power of the GF6800 that makes it possible, but the pixel shader capabilities of the card, but as already mentioned, GF6800 isn’t the only card capable of displaying this effect. And the result gets even better than your example, because it generates bumps on a per pixel basis as apposed to per vertex basis as in your example. Originally posted by SN-AVATARYou're so lost dude, Displacement mapping don’t do ANYTHING to the polygons, it’s just an effect making it LOOK like there are more polygons, and therefore you can make a brick wall or a rocky surface with only 2 polygons (which are minimum for a flat surface).
Shader 3.0 Download For Pc Games
Basically it makes a flat surface look bumpy.It's kind of an evolution to bump mapping, making you see the bumps from the side and make the bumps self shadowing, you are comparing the new displacement mapping techniques with old ones in CG programs. And it’s just not the same.The whole point is to make more detailed scenes without too much performance impact. It does make it easier to make a brick wall for example, but that’s not the main reason it is used.What you describe it as, is no more then height mapping, as used to make the terrains in for example UT2003/4.
Pixel Shader 3.0 Nvidia
So it’s not the raw rendering power of the GF6800 that makes it possible, but the pixel shader capabilities of the card, but as already mentioned, GF6800 isn’t the only card capable of displaying this effect. And the result gets even better than your example, because it generates bumps on a per pixel basis as apposed to per vertex basis as in your example.
Shader Models are components used to help render graphics sent from the CPU to the graphic card. Originally posted by:I have also ran into the same issue but with my laptop. I looked it up on the internet and found that my video graphics are not up to snuff as they are intergrated graphics (on the motherboard).
I use my desktop computer with a Nvidia GeForce 240GT because I like to play high graphic games. You might have to look for an upgrade for your computer. The latest graphics cards to have the Shader model 3.0 or higher. I hope this helps. You are commenting on something thats over 3 years old.you know this right?