The ghosting perception of TAA is really affected by the quality of the TV.
TV with lot of remanence have more ghosting.
Image Quality and Framebuffer Analysis for Available/release build Games *Read the first post*
- Thread starter Farid
- Start date
The ghosting perception of TAA is really affected by the quality of the TV.
TV with lot of remanence have more ghosting.
Looks like 1024 there.
The first trailer indicated 1152x720, and the new shots aren't different.
Right, and these captures show this still it seems.
Hey Guys,
Here's another comparison from GameBlurb, this time it is Spider-Man: Edge of Time. Looks quite close.
GameBlurb: Versus - Spider-Man: Edge of Time Comparison [720p/HDMI]
Here's another comparison from GameBlurb, this time it is Spider-Man: Edge of Time. Looks quite close.
GameBlurb: Versus - Spider-Man: Edge of Time Comparison [720p/HDMI]
I'm curious to see if they improved the resolution for Crysis on the ps3. Didn't Al already confirm it was still 1152 x 720 on the 360? I see no reason why they would change that.
Also is there a link about FXAA on Crysis 2? I always thought it was a custom edge detect AA while Brink was the first game to feature FXAA.
The ghosting was the result of TAA, plus there was supposed to be a custom edge AA for up close objects IIRC.
You want him to post something concrete for what we can't see?
I kid I kid.
I don't know any term to call this post custom AA effect, so I'm calling it 'ghosting'
http://xoticpcforums.com/showthread.php?t=11253
not the same article saw earlier but it mention the same thing except for Brink
"The FXAA in FEAR3 is based on FXAA 1. FXAA 1 was also used in Duke Nukem Forever, Age of Conan, Crysis 2 (with their temporal 2x supersampling) and Brink."
Didn't DF also listed C2 as one of the FXAA1 game. It was using FXAA1 + TAA for distance objects. Guess FXAA1 wasn't as effective as the newer version when it comes to sub pixel level stuffs. PS3 might be using it still on newer games like driver 3, DE3 effect result in the lack of AA on some sub pixel aliasing.
not the same article saw earlier but it mention the same thing except for Brink
"The FXAA in FEAR3 is based on FXAA 1. FXAA 1 was also used in Duke Nukem Forever, Age of Conan, Crysis 2 (with their temporal 2x supersampling) and Brink."
Didn't DF also listed C2 as one of the FXAA1 game. It was using FXAA1 + TAA for distance objects. Guess FXAA1 wasn't as effective as the newer version when it comes to sub pixel level stuffs. PS3 might be using it still on newer games like driver 3, DE3 effect result in the lack of AA on some sub pixel aliasing.
Hey thanks for the link.
I wonder if they are just mistaken with FXAA in Crysis 2. I mean, isn't any post process AA solution (MLAA, FXAA, DLAA, etc) nothing more than an intelligent edge detect blur? Looking back at the DF article, here's what they say: "temporal anti-aliasing is being used on objects a set distance away from the camera, with an edge-detect/blur mechanism in play for elements close to the player."
Is it even possible to restrict FXAA to close objects while ignoring objects further out? I thought the processing pass ran over the entire screen, which sometimes resulted in slight blur with text and such. If they could restrict the depth of FXAA, I still wonder why they wouldn't use it for the entire scene and ditch TAA entirely.
I could be way off here, but I don't think FXAA is used in C2. It's been a while since I played it, but I also don't remember seeing the same characteristics of FXAA as I saw in DE:HR on the 360.
Another thing I question in that article is how they mention FXAA was designed for the 360. I thought both FXAA2 and the improved FXAA3 were designed for the 360 and ps3. According to DF, Driver 3 does use FXAA on both consoles.
Yep.
One could use the depth buffer for edge detection parameters, though there are storage and performance costs associated with that apart from using the colour frame info.
The only game I recall using FXAA on the ps3 is Driver and I don't remember any mention of issues with sub-pixel aliasing in that title. Any other game I could think of actually uses MLAA (or some variant) on the PS3 which may explain the lack of sub-pixel coverage.
I could be wrong, but if C2 was using FXAA, TAA shouldn't be needed for sub-pixel aliasing.
Cool thanks Al. Makes sense on using the depth buffer and how it'll occur a cost.
This just leads me to further believe FXAA isn't being used in C2 when they were already running into memory and processing issues.
About dead rising 2: off record ps3 version: http://www.computerandvideogames.com/video/1211725953001/dead-rising-2-off-the-record-first-15-minutes/?mode=hd
any changement to the previous?
any changement to the previous?
Yes you can restrict it simply by drawing the full screen quad with z-buffering active ("less" depth comparison mode). This is very cost effective (z-buffering skips pixels completely that do not pass the depth test). Also console games tend to run FXAA filter before applying UI on top of the rendered image, so it doesn't blur the UI in either case.
This is assuming they can afford the cost on the GPU.
Cool thanks sebbbi. That may also explain why I saw jaggies in the UI for BF3,
Hey All,
The guys over at GameBlurb have just posted their screenshot comparison of Batman: Arkham City. They look pretty similar visually.
GameBlurb: Versus - Batman: Arkham City Screenshot Comparison [720p/HDMI]
The guys over at GameBlurb have just posted their screenshot comparison of Batman: Arkham City. They look pretty similar visually.
GameBlurb: Versus - Batman: Arkham City Screenshot Comparison [720p/HDMI]
Not one of which observations pertain to image quality and framebuffer analysis which is entirely focussed on rendering resolutions, buffer formats (HDR), and AA amounts.
Cross platform tech discussion belongs here.
Similar threads
