My ATI Results and Observations

[Kosta 122]

OK now that I have results, I am reopening this topic.

 

First of all, I wish I had a BMP of your full sized shot if possible? Can you email me that to Word Not Allowed@avsim.com please? I can then put a non jpeg-ed picture up. Still, your shot pretty clearly shows the aliased line.

 

On with the shots:

 

This is yours:

 

 

Now, this is mine with 2xSGSS + 8xS:

 

 

So I am already here killing the ATI by 10%.

 

On to quality comparison:

yours:

 

mine with 2xSGSS:

 

 

As you can see on your there are distinctive white and black points on the line, showing how the line aliases as going down.

On mine it's split into many little "parts", making the line visually from far away more attractive.

 

Now, lets give Nvidia a chance with lower settings:

 

 

This is what it produces on the line:

 

 

Notice much higher jaggies on both wing and gear line.

Very much comparable with the ATI in term how the line looks like (no matter if the line that goes to the wing or to the gear).

 

And now, that out of the way, what happens with the cloudtest on Nvidia GTX580 in comparison to ATI when ran with your settings:

 

 

I hope this clarifys anything you were doubting!

[Bobu 0]

Thank you very much for doing this Srdan. In my original post one of the possible reasons I hypothesized that I was seeing great performance was that I was not seeing what others with high end systems were seeing. Hence the creation of my video, and request for similar from an nVidia user. After looking at your screenshots, I can see that there is a difference.

 

The whole reason for my post was to understand first, were my visuals comparable with what others were seeing, and if so, why. Maybe my settings or combination of hardware could benefit others. On that note, when I look at your results and compare it with mine, i do see differences, but they appear somewhat subtle. Are there differences? Yes. Are they game changing? I am not sure. I guess each person would have to make that call.

 

My last question on this topic is related to expectations. Are you at all surprised by the performance I am getting, or is the difference between a 5850 and higher end nVidia cards within line? Because based on what I see in the comparison shots, it just doesn't seem like a massive improvement between a 2-3 year old card, and today's hardware. Or maybe a better way to look at it is that nVidia is the better choice for FSX, but the gap isn't that massive. If you have an ATI card, and can live with slightly lower visuals, maybe for FSX it is an acceptable choice.

 

Maybe if we had our systems side-by-side these tests would show more ground-breaking differences. Thanks again for doing this Srdan. It at least answers the main question, which is nVidia users do see a better visual than I do, which answers why their performance is not far greater at much higher AA. My other hypothesis was that I was comparing apples to oranges, and this is also confirmed.

 

In any case, if anyone wants my cfg file, please PM me. I also have the shader mod installed.

[Bobu 0]

Okay I was looking at the images on my iPad and missed the FPS counter in the top right corner. So your 580 gets double the fps of mine when using similar settings, and a little better with the higher AA. The only thing i don't know is how good your visuals look at my AA settings. I would imagine that since you are used to the higher AA you would really notice the jaggies and dancing lines. But I wonder if I would.

 

Maybe I will pick up a 580 and do the test. You have me wondering what I am missing. I would also like to know if I find the jaggies unacceptable on the 580 with the same AA settings I run on the 5850. Would you expect the same AA settings to provide different results? Or is 4X AA 4X AA regardless of card mfg?

[Kosta 122]

i do see differences, but they appear somewhat subtle. Are there differences? Yes. Are they game changing? I am not sure.

 

That is for each person to judge. For me they are a must. For you maybe not. Something I've said 5 times in this thread already.

 

or is the difference between a 5850 and higher end nVidia cards within line?

 

The difference is in line, but not between 5850 and higher nvidia cards, rather ATI and Nvidia in general. Something everyone on this thread keep yelling. Why? Because you went ahead and said something in lines that ATI is as good as Nvidia in FSX... no hard feelings man from my side :)

 

Or maybe a better way to look at it is that nVidia is the better choice for FSX, but the gap isn't that massive.

 

Gap is not massive? I mean, OK, your card is older, but I bet even newer ATI cards can't do that (last shot).

 

If you have an ATI card, and can live with slightly lower visuals, maybe for FSX it is an acceptable choice.

 

 

Why would you settle for less if you can have more for the same amount of money? If you already own ATI, sure, you have to settle with what you have. But then, what is the point of this discussion?

 

The only thing i don't know is how good your visuals look at my AA settings.

 

Multisampling is much a standard setting on both ATI and Nvidia, so they are comparable.

 

Would you expect the same AA settings to provide different results?

 

Not by much. As already said, MS is a standard in the industry. Both cards will make same pic quality with 4x MS. Only question is what is your supersampling really doing?

 

Can you test with 4x MS but without supersampling?

[Bobu 0]

Thanks again for the info. If AA is pretty much a standard, then I should see the same visual results on a higher end nVidia card with the 4X AA settings. If this is the case I am going to be really happy, because I should be getting amazing fps with FSX. Or...I am going to see jaggies and dancing lines, which puts this entire discussion into focus.

 

Just remember my point of this discussion was that I was getting extremely fluid and jaggie/tear-free gaming with very acceptable results, on an ATI, which seemed well out of the norm. If the nVidia gives me similar visual results at 4XAA (at much higher fps) then case closed.

 

So...based on what Ben said it appears that the 660TI is the card of choice. I will probaby pick it up and redo the tests. Could you leave the thread open so I can report back the results?

[Bobu 0]

And one of the points of the thread was trying to determine the source of the results I am seeing. Everyone was harping about AA and that it was the difference. I was told to run my settings with AA maxed, and even then it wasn't fair because ATI doesn't have all the AA settings. And my reply was always the same. I am not seeing jaggies, so why would I kill my performance running higher AA than I needed? That was why I posted the videos, so others could explain to me what I am missing. You gave me the closest explanation.

 

Telling me something doesn't make it so. Showing me does. I am a show me kind of person. And 95% of the comments I received were lacking of any substance, other than repeating what others have said.

 

I've been gaming a long time. I know what jaggies and dancing lines look like. Of course, we all crank our AA settings up to the point we reach that visual/performance balance. And for some reason, that point requires very little AA on my system.

 

Again, I will redo the test with the 660TI and see how it goes. I am really hoping you guys are right, and nVidia at 4XAA provides the same visuals with much greater performance. But what if it doesn't?

 

Only one way to know for sure!

 

 

[Ben Cap 89]

660TI is the card of choice.

 

Yes indeed! :Big Grin:

[Bobu 0]

Thanks Ben. I had a feeling you would give it the thumbs up.

 

BTW, i am finding plenty of discussion on the benefits of higher resolution vs. AA. Thinking about jaggies and dancing lines, it seems like they occur because the resolution of the monitor isn't high enough, so AA is needed to smooth it out. Many posts address this on Google, so I am thinking I can get away with less AA due to running 1920x1200, with less noticeable jaggies. If that is the case, the 660TI should provide similar results visually at the same AA.

 

Ben, would 2GB do or should I opt for a 3GB version? I don't see many options fir those on newegg, though.

 

 

[Ben Cap 89]

 

Ben, would 2GB do or should I opt for a 3GB version? I don't see many options fir those on newegg, though.

 

As long as you aren't running 3 monitors, 2GB is plenty.

[Bobu 0]

As long as you aren't running 3 monitors, 2GB is plenty.

 

Perfect. Only running one. Thanks Ben, and sorry for the attitude on this topic.

[Kosta 122]

I will probaby pick it up and redo the tests. Could you leave the thread open so I can report back the results?

 

Sure. Great thing. Pick it up and tell us how it goes!

[Bobu 0]

Sure. Great thing. Pick it up and tell us how it goes!

 

Will do! Thanks Srdan.

[Ben Cap 89]

 

 

Perfect. Only running one. Thanks Ben, and sorry for the attitude on this topic.

 

It was a mis-interpretation and an over reaction on my part...

[Bobu 0]

It was a mis-interpretation and an over reaction on my part...

 

No worries Ben. Always hard to gauge "tone" on a forum. All good.

[Clutch Cargo 384]

Yep, went through this debate with myself last year... AMD or Nvidia... which way should I go????? I even bought an AMD 6970 and a GTX 580 to make my own comparisons. I liked Eyefinity also compared to Surround so I was hoping AMD would win out.

 

Where I was getting some very good outputs and visuals with my AMD the difference was in the details. It took a little more tweaking with the AMD card and I still just could not get as good as results. I wasn't as scientific as Word Not Allowed but I could see with my own eyes what I liked better, what performed a little better. But what finally sold me was I did some very hi-res photoreal scenery creation at 15cm. I could see the detail better in FSX with my 580 than I could with my 6970.

 

In fact, I just bought a 2nd 580 so I can run my three 27" in Surround. But I am still an AMD fan. Just finished building an AMD 6-core to match up with that 6970 to run other programs outside of FSX, also as my multi-media system and also to serve as a back-up system.

 

Clutch