Performance Is HALF Of FSX:SE

[VeryBumpy 272]

The near double frame rate difference is massive for such similar settings. I agree there is some large underlying cause for such a difference. Please let us know what you find out.

[hesynergy 67]

If I were to load up ORBX scenery, and fly over it in the NGX with traffic, weather, and REX textures using FSX:SE I'd drop from 100+ FPS down to 30-40 FPS. That's doable.

 

Considering my performance in P3D is half that, I'd be dropping down as low as 15 FPS. Lower in bigger cities. No, I'm not happy with that. And no, that's not massive FPS. I'm not about to start shooting slide-show approaches in a $200 sim that behaves virtually identical to FSX but runs at half the speed. Obviously something has to be wrong with my set up somewhere. I'm sure there's tons of people out there running the sim on high settings without issues. This would be an extremely common complaint if every i7/900 series GPU owner started feeling like they were in 2007 again.

Just an old farts query...could it be that we Caint see the king is Nekit??

 

Chas

[nooon 21]

Thought I'd try making the switch to Prepar3D Pro since the FSL A320 Pro will be a Prepar3D exclusive, and I'm appalled by the performance I'm seeing given the same exact settings...

 

I'm gonna go ahead and assume I've done something terribly wrong in the set up process. Can anyone think of what that might be? Linked below are in-game screenshots I've taken at KBFI, in-game sliders matched as closely as possible (with the exception of hardware tessellation), fair weather, system time, and the default Maule aircraft. No mods installed on either sim. P3D averages 65 FPS, FSX:SE averages 120 FPS.

 

http://imgur.com/a/gfkIv

 

Make sure Dynamic reflections are off.

 

I'd try first old nvidia driver 353.62 as a starting point.

Then try to maximize tesselation slider, then minimize to zero : which one gives you the best FPS reading ?

If nothing improves, try disabling volumetric fog. Did FPS improve ?

[pmanhart 176]

You should be getting great performance in P3D. I never go below 25 FPS in Seattle with almost exactly same hardware specs, with almost all graphic settings maxed out (shadows and dynamic reflections included).

 

In fact, I was flying around Orbx's YMML last night in the F1 B200 (G1000 is a FPS killer) and only dipped below 30 FPS when I upped the weather.

 

Do not assume FSX-P3D settings are the same. Start with the default P3D settings and then work your way up from there. The sims are so far apart now in how they use your hardware and what the graphics settings actually do that you are doing yourself a disservice by trying to set them up similarly.

 

Are you using some super-magical FPS killing AA settings?

[AviatorMoser 353]

Yeah, watch out for antialiasing in heavy cloud conditions. P3D doesn't do those two particularly well simultaneously.

 

I've had similar experiences where the P3D performance would be about half of that in FSX:SE. It really had nothing to do with sliders and even image quality. It was just P3D's poor performance in heavy overcast when AA was enabled.

 

Couple of comparison (older) images below.

 

This was a slideshow in P3D:

 

And this was butter smooth in FSX:SE.

 

 

In these tests, it didn't really matter what the sliders were set to in P3D. The antialiasing was just drowning the GPU irregardless. I could have probably burnt my coffee on my GPU that day.

 

Anyways, my experience with P3D in lighter cloud conditions were quite pleasant. Good solid performance.

[ColeThePilot 0]

Do not assume FSX-P3D settings are the same. Start with the default P3D settings and then work your way up from there. The sims are so far apart now in how they use your hardware and what the graphics settings actually do that you are doing yourself a disservice by trying to set them up similarly.

 

Are you using some super-magical FPS killing AA settings?

I was only trying to set them up similarly because I could tell after a fresh install that my performance shouldn't have been so low. That's when I started trying to set them up with the same scenery complexity, mesh resolution etc.

MSAA is off, the in game AA setting was turned up a bit. I don't believe Nvidia control panel had changed anything by itself. I hadn't messed with that yet.

 

The only in game setting I've found (so far) that gets me up near FSX's performance is the scenery complexity, which when turned to its lowest setting gets me up to about 90-100 frames.

 

I'll post my settings later tonight, as well as my CFG file. I'm leaning towards something CPU/sim complexity related, since hardware tessellation being off doesn't make a noticeable difference, and in game AA/AF settings do not make a noticeable difference either.

 

Though I could be completely wrong. AviatorMoser - I'll check all of that tonight and report back with those screenshots and CFG file.

[tf51d 284]

I've seen reports of poor performance in P3D with systems with hyperthreading on, Try turning it off if it's on in your system.

[Bert Pieke 7,584]

The initial comparison is extremely misleading - who cares if you get 60 or 120 fps..

 

Set the sim to 30 fps and fly it into the areas you are concerned about.

 

If it is smooth, you are good - if it gets jerky, you have a problem and will have to reduce some settings.

I've seen reports of poor performance in P3D with systems with hyperthreading on, Try turning it off if it's on in your system.

 

Really..?

[Afterburner 374]

Here is what I can tell you from my experience:

 

Let's suppose that you run identical settings and that you initially have all shadows deactivated in both sims.

If you monitor the FPS immediately after loading the simulator, FSX:SE will display a considerably higher frame rate than P3D. However, if you switch views and start moving the view angle from the spot view, the FPS in FSX will drop and move closer to P3D (though it will be still higher).

 

On my system, as soon as I approach a more complex airport, the FPS in FSX drop even more compared to the initial loadup (sometimes by over 50%), while P3D maintains a more consistent frame rate during that stage. So I don't think that your expectation that P3D's FPS will drop by half during approach will turn out to be true.

 

Now here is where you gain from P3D: As soon as you start enabling all kinds of shadows (aircraft, VC, autogen, scenery objects, etc.) and water effects, the FPS in FSX will drop dramatically, because these features are very CPU-heavy (even under DX10), while your GPU is bored to death. This gets even worse with CPU-intensive airplanes (PMDG). In P3D, it's your GPU that is taxed, which will give your CPU still enough breathing room and leave you with more FPS despite better visual quality.

 

So in a nutshell: If you keep the eye candy deactivated, you are able to achieve higher FPS in FSX than in P3D. However, these FPS are not very consistent, since they can drop heavily during an approach. As soon as you activate shadows and water effects, P3D will win big over FSX in terms of FPS.

[ColeThePilot 0]

The initial comparison is extremely misleading - who cares if you get 60 or 120 fps..

 

Set the sim to 30 fps and fly it into the areas you are concerned about.

 

If it is smooth, you are good - if it gets jerky, you have a problem and will have to reduce some settings.

 

Umm...I care? For me, flying around at 12 FPS isn't acceptable. The reason I'm able to get acceptable frame rates in FSX is what we're trying to address here.

 

It's not misleading at all. Telling me I'll maintain 30/ "smooth" frame rates with GSX, addon scenery, PMDG aircraft, traffic, and complex weather scenarios when FSX just barely handles that and is performing better for me is what's misleading.

 

 

On my system, as soon as I approach a more complex airport, the FPS in FSX drop even more compared to the initial loadup (sometimes by over 50%), while P3D maintains a more consistent frame rate during that stage. So I don't think that your expectation that P3D's FPS will drop by half during approach will turn out to be true.

You may be on to something with this. I'll load up some addons and see if my assumption turns out to be true. Maybe it won't scale the way I'm thinking. 

[ColeThePilot 0]

This is really interesting you guys. Afterburner was right.

 

The sliders/P3D providing higher end scenery with sliders turned up wasn't the culprit. It's about how the two sims manage the FPS degradation as addons become a factor. 

 

I installed the NGX into both sims, as well as Orbx FTX scenery, and selected the thunderstorm weather preset. FSX:SE's frame rates went from approx. 120 FPS, to 45 FPS. P3D's frame rates went from 60, to 50 FPS!

 

It seems Prepar3D doesn't allow it's frame rates to reach the same dramatic levels of FSX, but the frame rates also seem to drop very little when addons become a factor.

 

Here is the new comparison screenshots. Top image is FSX.

 

I'm very happy about this. It's true the two sims cannot be compared 1:1, and I was wrong in my assumption that performance would be severely impacted once addons were installed into Prepar3D the same way they're impacted in FSX. Sorry for doubting.

[magnetite 38]

The reason I'm able to get acceptable frame rates in FSX is what we're trying to address here.

 

LM moved a lot of the stuff which used to be done with FSX on the CPU onto the GPU. 

[spesimen 135]

 

 

I'm very happy about this. It's true the two sims cannot be compared 1:1, and I was wrong in my assumption that performance would be severely impacted once addons were installed into Prepar3D the same way they're impacted in FSX. Sorry for doubting.

 

haha i don't think you need to apologize, it sounds like you had a hypothesis and you tested it.

i think that's actually pretty interesting result that the scaling of performance varies a lot between the sims but it seems like a reasonable result from whatever optizimations LM has been doing..

thanks for posting your results, i have been holding off from going to p3d specifically because i haven't really been convinced that i'm going to see enough of a performance upgrade to justify the costs. so it is useful for me to hear about it especially from people who aren't satisfied with 15 or 30. 

 

cheers,

-andy crosby

[basementflyguy 156]

This is really interesting you guys. Afterburner was right.

 

The sliders/P3D providing higher end scenery with sliders turned up wasn't the culprit. It's about how the two sims manage the FPS degradation as addons become a factor. 

 

I installed the NGX into both sims, as well as Orbx FTX scenery, and selected the thunderstorm weather preset. FSX:SE's frame rates went from approx. 120 FPS, to 45 FPS. P3D's frame rates went from 60, to 50 FPS!

 

It seems Prepar3D doesn't allow it's frame rates to reach the same dramatic levels of FSX, but the frame rates also seem to drop very little when addons become a factor.

 

Here is the new comparison screenshots. Top image is FSX.

 

I'm very happy about this. It's true the two sims cannot be compared 1:1, and I was wrong in my assumption that performance would be severely impacted once addons were installed into Prepar3D the same way they're impacted in FSX. Sorry for doubting.

I want to give you credit for being open-minded while folks tried to help, and really gracious in the reply above. It's FAR to common in Internet forums for people to entrench and become jerks, and this was class.

[Afterburner 374]

I installed the NGX into both sims, as well as Orbx FTX scenery, and selected the thunderstorm weather preset. FSX:SE's frame rates went from approx. 120 FPS, to 45 FPS. P3D's frame rates went from 60, to 50 FPS!

 

It seems Prepar3D doesn't allow it's frame rates to reach the same dramatic levels of FSX, but the frame rates also seem to drop very little when addons become a factor.

 

Here is the new comparison screenshots. Top image is FSX.

 

I am glad that you are happy with the results!  As you can see, P3D may not deliver the extreme, initial performance of FSX when no heavy add-ons are used and/or the eye candy is turned off, but as soon as you add these elements, P3D has a better capacity to withstand the intensive load when it comes to performance.

 

I am sure that if you compare the FPS between FSX:SE and P3D using PMDG 737 on approach (especially with the eye candy activated), you will find an even greater difference in performance between FSX:SE and P3D in favor of the latter. Give it a try!  Personally, I prefer having a more consistent performance at all flight stages over experiencing a jump between two extremes.