<Large Images>: Beta 1 vs Beta 2 terrain screenshot comparison

#1 have not seen that bug yet on my Core i7 920, GTX 280 182.50 drivers drivers, dx sdk march 2009, vista 64 business

#2 I've seen that on my computer on a few terrains so far.

AMD Opteron 165
2GB RAM
Geforce 8800GTX
dx sdk march 2009
185.85 drivers
Vista Ultimate x64

#3 Worth mentioning that I've turned up the shadow range a bit so I can get a wider picture of the terrain errors. The errors still exist at default shadow range, but obviously you then have to be close enough to the terrain for shadows to render at all to see them. You may need to get pretty close to the terrain and fly around a bit before you really start seeing the shadow errors. Increasing the terrain size will definitely make them appear.

Intel Q9650
4 GB RAM
GeForce 285 GTX
All drivers up to date -- just checked today
XP x64

#4 Same problems here (warrior camp)

Have tested it with 8800GTX and GTX285
182. and 185. (newest) driver
XP 32
Intel Q 9450
4GB

#5 I'll go ahead and re-post my images here as well.

These pictures are BOTH in Beta 1, the only thing I changed was the exe itself. So, all settings are identical, all art files are identical, etc. Only the exe itself is different.

Beta 1 exe:

Beta 2 exe:

#6 There is something new going on that occur sometime after beta 1. I'm deep in terrain work at the moment, but i'll be looking at this soon.

#7 Tom.. thank-you for that.. all we were looking for was a response that
meant our observations were being taken on-board :)

If there is anything I / we can help with, just ask :)

#8 I was having this problem in Beta 2 and am still having it in Beta 3.

I think it's just the lighting, but it gives the terrain a grid effect in certain areas. Odd, harsh shadows — similar to the images above. Can happen right away or only after I sculpt the terrain.

Other folks still experiencing it?

#9 yes

#10 yes

#11 For anyone still experiencing this in beta 3:

Open terrain/terrCell.cpp and look for the TerrCell::cullCells() function:

// If we have a VB and no children then just add 
   // ourselves to the results and return.
   if ( mVertexBuffer.isValid() && !mChildren[0]  )               
   {
      outCells->push_back( this );
      return;
   }

   const F32 screenError = state->isShadowPass() ? 0 : // CHANGED
                              mTerrain->getScreenError(); // CHANGED

   for ( U32 i = 0; i < 4; i++ )
   {

Let me know if this clears up your issues.

If not then also try reducing the terrain screen error in the inspector down to 0.

Finally... if that doesn't work either... try disabling the Sun shadow via the checkbox in the inspector. That should remove the artifacts... if it doesn't let me know.

#12 Is this still a bug for anyone in 1.0.1?