• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Possible huge ATI tessellation performance boost with 10.5 drivers

I'm no expert, but to try and address a few mis-conceptions - even by those that should know better:

TruForm was fairly simplistic in that it smoothed the boxy edges of a mesh, with varied results generally making things look bloated. It subdivides the mesh (also known as tessellation) but it is NOT what we refer to as tessellation now (see below) it does not add any additional fine detail to the polygons just makes the overall shape look less blocky.

Bump mapping tries to give the texels (pixels) on a polygon a 3D look - but in its normal form lacks parallax and only presents the illusion of depth from a small range of angles, it does not add extra polygons.

What we call tessellation now is actually badly named. I'm not sure what the correct term for it is off the top of my head. Tessellation is just one component of the entire process which is to sub-divide a polygon (tessellate) into lots of smaller ones, smooth the profile of the original shape and then extrude the vertices of the smaller polygons away from the original polygon based on the same type of data that is used to construct one form of bump mapping. So you are both smoothing the mesh so it doesn't look so blocky and adding extra detail to the polygons in a way that has parallax when viewed from any angle.

Sorry I'm not capable of explaining it any more simply than that :S
 
Last edited:
Good Rroff, I think I get it:) Could you achieve the same (more detail) using normal mapping? Or is that for a different purpose alltogeather?

Thanks
 
Normal maps/bump maps can potentially achieve higher detail than even extreme tessellation, but have the same issue with only looking convincing from a narrow range of angles and lack any kind of parallax unless you use displacement/occlusion mapping techniques.
 
Truform = tessellation read the whitepaper I referenced above it basically says exactly the same thing Roff said. :)
 
Not quite sure what your asking. There are techniques (polybump) for generating normal maps from high detail objects to apply to a low detail version of the mesh.
 
Biffa said:
Truform = tessellation read the whitepaper I referenced above it basically says exactly the same thing Roff said. :)
Click to expand...

Yes truform tessellates the mesh and then smooths it but its not the same as whats called "tessellation" nowadays - which is (silly ATI) rather badly named as its only a component of the whole process... I prefer what nVidia have done and referenced it as a polymorph engine. Gonna have to look up what the umbrella term for the whole process is really called.
 
I don't know anything about tessselation other then what I read on these forums, but does anyone remembe Dave Perry from Shiny pimping his tessellation engine on the game Messiah (pretty decent game I thought actually). He basically said as hardware got more powerful the game would look better, and better, so the user could re-experience the game at some point the way the developers meant it to be played :) That was the 1st time I heard the word tessellation. Anyone remember this game ?
 
IIRC messiah uses a description system, models were made at insanely high detail and then compressed with a format that mathematically "described" what they should look like, so you could at runtime rebuild the model at any arbitary detail level from a simple cube all the way upto as close as you could get to the original high res mesh sans detail that couldn't be rebuilt from the "description".

This technique is used in a few things like 4k/64k demos.
 
Last edited:
Last edited:
Not sure if this has been posted, from catalystmaker twitter feed in response to this : "those forum posts are full of misinformation - dont take them seriouslly"
 
Well pretty much confirmed from the horses mouth I guess... hope it wasn't charlie who wrote the original piece - tho it seems to be atributed to oc.com.
 
Rroff said:
Well pretty much confirmed from the horses mouth I guess... hope it wasn't charlie who wrote the original piece - tho it seems to be atributed to oc.com.
Click to expand...

Well , with the source not confirmed and as it sounds SO like summat he would write, I'm putting it down to Mr HugelyInnacurate himself until proved otherwise.....
 
The only thing I've known Charlie be entirely wrong about is the tessellation stuff back along - and the quote has the same style of mis-conception which makes me wonder but its hardly conclusive.

Why Charlie isn't often wrong tho is partly due to him often hedging his bets with something thats just as true even if its not the actual outcome, out of the possible outcomes, that happens, bit hard to explain. To be fair tho hes put out more correct information by a long shot than anyone else by comparison.
 
You must log in or register to reply here.
Back
Top Bottom