CSAA, Transparency AA (Multisampling), & Texture Sharpening (MIP LOD Bias) (w/pics)

I think my GFX card would roll-over and die if I attempted 8xSGSS and 16xQ

 

Why does it have to be an in-game setting? Why not just set the AA from Inspector? I'm pretty sure there is no difference (as long as the level you choose corresponds to the same AA amount in Inspector, for eg. I'm pretty sure with my 780 Ti that level 5 corresponds to 16xCSAA. I could be wrong though.) If you do want to set the AA from the game, while setting transparency AA from Inspector, then it should still work. If not, try changing the AA mode from application setting to enhance.

 

If you read any of what I wrote above, that's what I have been trying to do to no avail. No combination of settings works. Others are having the same or similar or sometimes opposite, but equally frustrating experiences. SS is not being applied under any normal use of NCP or Inspector, with AA set within rF2 or by NCP or by Inspector. SGSS works fine, but I would like a less taxing option for performance reasons, especially since so many tracks and cars in rF2 are not optimized. Applying 2xSS to any normal mode of AA (rF2 or NCP) shouldn't be challenging or require endless fiddling and frustration...but apparently I expect too much.

 

It is called 4 xAA lool

and it looks fine........... ?

I mean seriously, as example people like me with GTX670 ? Mucking around with more ? It is no wonder they can't lap hard and smooth. lol

You letting visual candy over rule your driving. wt ??

It's insane !


What does a still screenshot prove , online 20 cars moving is what matters, not static and not some baseline benchtest either.

 

Yes, why do we use win 7-8 and rFactor2, lets turn back time to 1987 , win 2.0 and Indianapolis 500, no need to benchtest ether.

So pathetic DD

 

What has what OS you use got anything to do with what I said.


The point is on average people are more likely to use more candy then less possibly adversely affecting smooth lapping.

It is pretty basic science.




Pathetic how ?

For some people to find they may get smoother racing and steering from dropping settings.
Maybe more consistent racing, maybe happier for it ?


That's a bad thing ? lool






P.S. Ari



You have this weird perception I think all anyone needs is my PC specs and a blurry 60Hz monitor. lol

I never implied that Ari ever, ..........you have dreamed that up by yourself.

I am saying your specific hardware in relationship to your specific settings.

If you can't get any smoother wheel with your preferred settings.......well simply disregard everything I said.....I mean isn't that obvious ?.

But if you have less powerful PC maybe, just maybe lowering settings will have some lapping smoother and suddenly they see a new lap record / Personal Best Time ...............and they will probably be stunned. lol



Tell me that makes no sense, i dare you.

 

Durge, give it up. You have made it clear multiple times that you think this all a waste of time. Fine. This thread is for people who do not think that way. You can just ignore us crazies, you know.

 

Sorry Marc, it's not that I'm not reading your posts, I just start to loose track of who said what since there are a fww different discussions going on here (TS, AA and CSAA, supersampling modes).

Ok, so when you say you're trying to use a non-sparse grid form of supersampling, are you trying to do this by still using supersampling (but not sparse grid) in the transparency supersampling options, or are you talking about setting the general antialiasing itself to a supersampling mode (for example 3x3, 8xS, etc.)?...

I'd like to help test and see what's going on...

 

Thanks. All I need to do is use Level x AA from rF2 and apply 2x or 4x SS in the NCP. Couldn't get simpler. However, it doesn't work. Details of what and why are in the previous many posts.

 

yes it makes no sense.

Everybody knows that 60hz vs 120hz is input lag vs no input lag, 60hz = 60fps and it don't mater what gfx settings you are using there is no improvement of lap times as long you have 60 fps, no stutters, tearing etc. when you are driving and same goes when 120hz = 120fps............ but of cause you don`t have any noticeable input lag with 120hz and 120 hz is not about candy, it is simply about to have better experience.

I can also read between the Lines that you are jealous to Spinelli, me etc. about the hardware we own, there is no need to this, I am not born with silver spoon in my mouth , I did hit the road from my home when I was only 14 years old and today 45 years later, I have what I need so simple is that.

 

Once I'm confident I've worked out exactly what's going on, I've seen some stuff today that has made me doubtful if I've really uncovered the full story yet.

 

This is what I'm thinking so far, but it's just pure speculation....

1. Like TechAde said, it obviously has something to do with different ways the track-creators create the tracks. Different material settings for different parts of the tracks, alpha-this, blend-that, haha I don't know, I'm not a graphics creator guy (just more into graphics tuning/settings).

2. This is the weird part though. The issue only happens with TS @ 0 (off). If this setting is just a MIP bias of 0, then it is just a step down from a MIP bias of +1 (TS @ 2), and a step up from a MIP bias of -1 (TS @ 3) (stating the obvious here - 0 is one way from +1, and is also one away from -1, nothing else to it). So, then why does this issue ONLY happen when TS is set to 0 (off), and no at any other setting. This to me implies that there is something particular to the, and only to the, TS @ 0 (off) setting. So, maybe the TS @ 0 (off) setting is more than just another MIP bias step. This is what we got to figure out, what is so special about this one setting, why is the graphics issue - when it does happen - only happening with TS @ 0 (off) and no other setting?....


P.S. I also find it odd that ISI calls TS @ 0 "off" instead of "0" (since everything else is a number (eg. +2, -1, -1, -2). It could be nothing at all, but it is inconsistent, no? Theoretically speaking, would there be a difference between setting a MIP bias to "0" and setting it "off"? Maybe "off" doesn't truly mean "0", but something else? Again, just speculation and questions, I'm not saying any of this is necessarily true.

 

CSAA, Transparency AA (Multisampling), & Texture Sharpening (MIP LOD Bias) (w...

I don't know whats going on with all those AA options, but ever since I switched from AMD to nVidia, I am strugling to get nice image without aliasing and I just can't. Tried all kind of combinations from some people here, something of my own too, also the closest thing recommended - equal amount if msaa and sgssa - and the lines on the road and objects further away like kerbs or crowdstands are always noticabely jagged.
With AMD was pretty simple, you had one option to choose, 4xSuperSample and that was it, the image was so nicely sharp and antialiased.

All those NV options are driving me nuts, don't know what to do anymore.

 

+1 !!!!!

But I hope it's just because something is preventing SS from working for the moment. It isn't working for me (if you have read the endless posts above). It should look OK once it is actually working.

 

Yup. With AMD just simply setting it to supersampling had even better IQ in ISI based sims than Nvidia with Sparse Grid Supersampling. I was with AMD for 10 or so years before moving to 780 Tis. Seriously, just set to 4x Supersampling (which gave 90% of the image quality of 8x Supersampling but at a much, much less performance cost) or set to 2x Supersampling if you need to, and you're done, simple as that.

Having said that, apparently Sparse Grid Supersampling has a slight bug which means you need to manually change the LOD bias, and for some games, also the SLI bits as explained here --> http://translate.google.com/translate?hl=en&sl=de&tl=en&u=http%3A%2F%2Fwww.forum-3dcenter.org%2Fvbulletin%2Fshowthread.php%3Ft%3D490867, that may be part of the issue as well but I'm not sure.

Maybe it's down to different methods of supersampling that each manufacturers use.

[TABLE="class: docutils"]
[TR="class: row-even"]
[TD]SSAA
[/TD]
[TD]Super-sampling Anti-aliasing
This is the type of AA that creates the best image quality. It is “pure” AA with no compromises. All the other AA technologies exist because SSAA is very slow.
SSAA works by calculating multiple samples for each actual pixel on the screen and then averaging those samples to determine the color for each pixel. The ratio of samples to pixels is often described by a number put before the acronym. For instance, 8xSSAA describes that each pixel on screen will be the avarage of 8 samples.
When there is no multi-sampling, the color of each pixel is determined based on a single sample from the center of the pixel. When multiple samples are calculated per pixel, those samples are distributed across the pixel in patterns designed to maximize the AA effect. In particular, this means that the samples are not evenly distributed and do not line up horizontally or vertically.
SSAA is slow because the pixel rendering stage of the rendering pipeline must do the same amount of work as if the screen resolution was many times the resolution of what it is. For instance, with a resolution of 1920x1080 and 8xSSAA, the hardware must calculate the same amount of pixels as if the resolution was 7680x2160.
[/TD]
[/TR]
[TR="class: row-odd"]
[TD]OGSSAA
[/TD]
[TD]Ordered Grid Super-sampling Anti-aliasing (OGSS)
Same as SSAA.
[/TD]
[/TR]
[TR="class: row-even"]
[TD]RGSSAA
[/TD]
[TD]Regular Grid Super-sampling Anti-aliasing (RGSS)
Same as SSAA.
Note: RGSSAA has two separate meanings. See the entry below.
[/TD]
[/TR]
[TR="class: row-odd"]
[TD]RGSSAA
[/TD]
[TD]Rotated Grid Super-sampling Anti-aliasing
Similar to SSAA. The difference is that instead of the sample grid being along horizontal and vertical axis, the grid is rotated to a fixed angle. This causes better anti-aliasing of edges that are almost horizontal or vertical.
Note: RGSSAA has two separate meanings. See the entry above.
[/TD]
[/TR]
[TR="class: row-even"]
[TD]SGSSAA
[/TD]
[TD]Sparse Grid Super-sampling Anti-aliasing
Similar to SSAA. This type of AA is the first in a series of AA types that represent a compromise designed to trade image quality for performance.
The sampling pattern of plain SSAA is called Regular Grid or Ordered Grid. It can be pictured as a grid that has higher resolution than the pixel grid. Where SSAA samples in each square in that grid, SGSSAA samples in only some of the squares.
[/TD]
[/TR]
[TR="class: row-odd"]
[TD]JGSSAA
[/TD]
[TD]Jittered Grid Super-sampling Anti-aliasing
Similar to SSAA. The difference is that sample locations are altered so that they no longer exactly follow the grid that plain SSAA uses. This helps hide regularly occuring artifacts.
[/TD]
[/TR]
[TR="class: row-even"]
[TD][/TD]
[TD]Stochastic Super-sampling Anti-aliasing
Similar to JGSSAA. The difference is that the sample positions are fully random instead of being just randomly offset from the grid as in JGSSAA. The white noise that this introduces helps hide rendering artifacts.
[/TD]
[/TR]
[/TABLE]

Source --> http://www.dahlsys.com/misc/antialias/

 

Glad to see that I'm not the only one who finds that AMD does a better job at IQ with rF2 than nVidia with "much less performance cost".

Spinelli, just curious, when you use SGSS and change the LOD bias in Inspector, do you notice any difference (for example -2 to +2)? With TS =0, do you notice a difference where negative values sharpen and positive values blur? If you do see a difference, I'm still confused as to why TS doesn't do the same.

Also, TS =0 still confuses me. If it's "off", shouldn't the code in rF2 simply skip any subroutines that modify LOD bias? On the other hand, if it this isn't true and it doesn't disable anything, does it set everything to zero?

 

Glad I'm not the only poor confused soul!

Sent from my SM-G900F using Tapatalk

 

"Everybody knows that 60hz vs 120hz is input lag vs no input lag, 60hz = 60fps and it don't mater what gfx settings you are using there is no improvement of lap times as long you have 60 fps, no stutters, tearing etc. when you are driving and same goes when 120hz = 120fps"

Exactly!

 

It is poorly named in the .JSON file description. Confusing. Read above for how it actually works.

 

Went back and re-read posts. Seems I missed the Spinelli post on the 17th. Spinelli confirms what I had observed - that LOD bias doesn't seem to affect all tracks the same way. Fences, for example, can look fine at one track with a specific bias setting and not so great at another with the same setting. Seems the difference (or blame) may fall upon track makers. I tend to use Mid-Ohio as my baseline, since I think it has good material settings.

I hope TechAde can shed more light on this if he has the time.

This thread is getting confusing because multiple topics are being discussed. People have spent a lot of time testing and reporting their findings. Maybe Spinelli, or someone else, will volunteer to consolidate it all? Or maybe this isn't possible yet?