I´m working on a terrain rendering using GPU-based clipmaps mentioned in GPU Gems.
In description of VBO an IBO is mentioned following
we can render the (x, y) geometries of all terrain blocks using a single read-only vertex buffer and index buffer, by letting the vertex shader scale and translate the block (x, y) geometry using a few parameters.
If I understand that correctly, than the entire terrain grid that is visible, should be stored in one VBO and IBO. Is that correct? I´m just asking because it sounds ´too much´?
Yes, it is correct! And I don’t see any problem with that.
But this algorithm enables only flat grid. It is very fast, but does not enable rendering real terrains of large scale. What kind of terrain are you trying to model?
Thank you for the reply,…well…I´m just learning. This method I have found was the best described (for my current knowledge).
Well, I don´t have any specific terrain yet, but in the material is mentioned 1024x768. Data set was 216000x93600 height map. There is mentioned that it compresses by factor of more than 100 (don´t understand yet) and therefore fits in 225MB.
They compressed residuals between pyramid-levels using 2D image encoder, and published a compression rate 60-100x. If you plan to develop a game-engine, then this algorithm suits you very well. But if you try to model a whole planet or a significant part of it, then … I don’t think so.
