1 registered members (AndrewAMD),
684
guests, and 5
spiders. |
Key:
Admin,
Global Mod,
Mod
|
|
|
Re: the epilogue of a success
[Re: txesmi]
#436269
01/22/14 17:48
01/22/14 17:48
|
Joined: Oct 2007
Posts: 5,210 İstanbul, Turkey
Quad
Senior Expert
|
Senior Expert
Joined: Oct 2007
Posts: 5,210
İstanbul, Turkey
|
Okay. Pretty cool.
But the problem with your method of anti-aliasing is that even though it looks better than no anti-aliasing in screenshots, it sometimes makes aliasing more noticeable in motion. It looks like the aliased lines are "walking" when you look from certain angles and slowly rotate the camera.
It also seems to me like SSAO(or some other effect) is 1 frame behind.
3333333333
|
|
|
Re: the epilogue of a success
[Re: Quad]
#436270
01/22/14 17:55
01/22/14 17:55
|
Joined: Jun 2009
Posts: 2,210 Bavaria, Germany
Kartoffel
Expert
|
Expert
Joined: Jun 2009
Posts: 2,210
Bavaria, Germany
|
looks really awesome, but it doesn't look like ssaa to me (wrong offsets / wrong downsampling?) edit: oh and just in case you didn't know: wrs doesn't prevent curious people like me from looking at your shaders (use pre-compiled shaders instead):
Last edited by Kartoffel; 01/22/14 18:09.
POTATO-MAN saves the day! - Random
|
|
|
Re: the epilogue of a success
[Re: Kartoffel]
#436310
01/23/14 09:30
01/23/14 09:30
|
Joined: Jun 2007
Posts: 1,337 Hiporope and its pain
txesmi
Serious User
|
Serious User
Joined: Jun 2007
Posts: 1,337
Hiporope and its pain
|
@Kartoffel You are totally right, sorry. I botched the supersampling xP now the supersampled image looks as expected
float2 texCoor = inTex.xy / 2.0;
texCoor += vecViewPort.zw * 0.25f;
finalColor.rgb = tex2D ( mtlSkin1Sampler, texCoor.xy ).rgb;
and this is the downsampled image
float2 offset[4] = { 1, 1, -1, 1, 1, -1, -1, -1 };
float2 texCoor = inTex.xy*2.0f + vecViewPort.zw*0.25f;
float2 fScale = vecViewPort.zw * 0.125f;
for ( int i=0; i<4; i+=1 )
{
float2 Offset = offset[i].xy * fScale;
tex.rgb += tex2D ( mtlSkin1Sampler, texCoor.xy+Offset.xy ).rgb * 0.25f;
}
|
|
|
|