Comments on: Papervision SSAO phase II

By: Edan KWAN

Edan KWAN — Thu, 25 Feb 2010 07:50:42 +0000

I just made one inspired by this.
http://edankwan.amiciforce.com/

I used your Depth Buffer Material, its brilliant! The performance of mine is really bad. I have already low down the 2200 width of the buffer into 256 which mean i just need to take care of the blue channel and i used 4 random points for SSAO detection for each loop.

The performance is still bad and the effect is not as good as yours…

By: Ben

Ben — Tue, 15 Dec 2009 11:57:18 +0000

@Alex thanks! Source is coming soon

By: Alex

Alex — Wed, 09 Dec 2009 15:30:03 +0000

Really amazing!
I would like to try it, so I wonder if you could provide the source code…

By: nicoptere

nicoptere — Tue, 29 Sep 2009 10:29:03 +0000

that’s rather impressive.

can’t get really how you find the final value but it works really well both for shading and blooming.
keep on

By: Ben

Ben — Mon, 28 Sep 2009 16:40:20 +0000

Thanks! Currently I am sampling 8 depths per pixel using pixel index to generate a sine value which then drives the positioning of the depth samples taken (poor mans random generation). I did contemplate using the 2nd PB input as a way of passing in a random LUT and rotating through those values, I’ll have to give that a try!

By: Ralph Hauwert

Ralph Hauwert — Mon, 28 Sep 2009 16:12:33 +0000

Looking good. I was working on exactly the same thing for FP10. A tip I might give you is to do the depth sampling using a pseudo random lookup for surrounding pixels….this would remove the static look of the noise, and might help on the overall look of things, when moving.

Great work!