Half Life 2: Episode Two / Orange Box Feedback

[bob saget]

Originally Posted by methimpikehoses

Ha! You know it!

thanks bob, I got the buggers

nice bro....and i will ignore the comments made by craing....

[crainger]

Also, my suggestion to you guys is to not respond to crainger. He will try to rile you, and he's not worth the trouble.

[bob saget]

Originally Posted by crainger

Also, my suggestion to you guys is to not respond to crainger. He will try to rile you, and he's not worth the trouble.

i thought that was camp0rz
oops.....
i think i just.......
will you be gentle? ?

Holy short game, batman! (ep 2)

[NarcissistZero]

Originally Posted by methimpikehoses

Holy short game, batman! (ep 2)

But oh so sweet.

[bob saget]

Question:

I am playing at 1680x1050. There are some strange settings for AA/AF in Ep.2.
What would be the best looking combo without installing mods or anything like that?

[Noriega]

Just got mine OB

[Sazar]

Originally Posted by bob saget

Question:

I am playing at 1680x1050. There are some strange settings for AA/AF in Ep.2.
What would be the best looking combo without installing mods or anything like that?

Not sure. At the moment I max everything and have motion blur full and HDR full. It still looks its age though

Personally, I would HIGHLY recommend getting the latest cinematic mod. It just makes the entire game (HL2, Ep1 and Ep2) look gorgeous and many of the new textures are pretty awesome.

[bob saget]

Any difference between these? Filtering I have set to 16x.

MSAA
CSAA
QCSAA

Originally Posted by bob saget

Any difference between these? Filtering I have set to 16x.

MSAA
CSAA
QCSAA

Yeah. Ok, from the top:

MSAA: standard programmable multisample algorithm w/w/o gamma correction
CSAA: coverage sample msaa. ex: 8x: uses 8 (NO, uses 4) z-samples (depth), ignores the extra 4 color samples (over 4xMSAA).
XQCSAA: uses all of the color bits. gives true msaa modes over 4x. I believe the distinction is mostly due to the fact that the ROP's have to use some kind of loopback for CSAA. ROP's can only do 4x single-cycle. It's all marketing at this point. 8x MSAA and 8xQ CSAA are one and the same.

correct me if I'm wrong.

edit: looks like CSAA modes still use the single cycle storage values for pixels; i.e z and color (says stencil also, but...what the hell are stencil values unless we're speaking of stencil volumes...i guess that must be it). so depth value is still at 4 samples for 8x CSAA. For Q modes, z/color/stencil sample # is at the appropriate level for the stated MSAA #. So 8xQ CSAA applies 8 coverage samples (# of intermediary steps applied to sub-pixel edges) using 8x scene color/z readings (base level of inputs used to determine the position of the fractional values).

So 8x CSAA basically samples each pixel 4 times to determine the position of the sub-pixel data offenders in image space, then applies 8 samples instead of 4 at these sites to give the impression of a smoother appearance (guesstimating how the edge should look based on 4 samples). However, if you have data that can't be represented by sampling the area 4 times, then you will have no benefit over regular ol' 4xMSAA. clearer?

edit2: p.s: I'm not trying to give the indication that any sort of edge detection is going on. it's not. samples are taken from the entire image space, unless the game provides a custom coverage mask. however, only sub-pixel infractors are blurred, because layering data over whole pixel representations yields no new information.

[bob saget]

The urine jar for sniper is awesome

[mullet]

I Finally got my ENGI HAT!!!!!!!!!!!!!!!!!!!!!!!!!! LIFE IS COMPLETE!!!!!!!!!!!!