It's a bit hollow saying "those results don't represent what FSX can do", when on this system and config you're seeing what it actually does do, and then with a comparison to FSX:SE, which performed considerably better.

 

Craig, I've very much appreciated the information you've shared, but...

 

I think you're actually missing Ethan's point, and it's a fair one.  I assume that one of the reasons you're sharing this information is to give others who've not taken the plunge yet a relative idea of how SE performs when compared to FSX, but when your FSX numbers are well below what many of us get in similar situations, it does beg the question as to whether the marked difference you're seeing is due more to FSX install/config issues on your machine than it is to SE improvements.  That's partly why I raised a virtual eyebrow earlier at your Duke performance issues, as the V2 Duke runs very well on my modest system.  As you correctly pointed out, this may be due to FSX's general preference for nVidia over AMD... or it may be something else haywire on the FSX side on your system, as opposed to something inherently broken in FSX that SE improves upon.

 

Please take what I've said in the spirit it's offered.  The information you and others are providing is valuable and much appreciated, and all that I'm reading taken in total does appear to be painting a very positive picture for the SE changes.  I've gathered enough information that I'd be jumping in as well were it not for some specific add-on compatibility issues which remain showstoppers, despite the fact that my FSX install has been a solid performer.

 

Scott

But that's the whole point though isn't it? (that you've gloriously missed). Yes the performance in the FSX screenshots is poor. But with the same hardware and software, with the FSX:SE installation, it is not poor. There lies the comparison!

 

It's a bit hollow saying "those results don't represent what FSX can do", when on this system and config you're seeing what it actually does do, and then with a comparison to FSX:SE, which performed considerably better.

 

It doesn't matter whether you're testing FSX and FSX:SE on a NASA super computer, or a Sinclair ZX81;  as long as you're testing both sims with the same conditions, then you get a reasonable comparison....  :rolleyes:

 

I changed all my FSX settings to be the same as yours, except  DX10 mode is enabled and my resolution is a more demanding 2560x1440. I am using antistropic filtering (16x via Nvidia Inspector), and Nvidia Inspector AA is set at 8x MSAA w/ 2xSGSS. I have a 2600k w/ a GTX780. I have a faster GPU, but I don’t believe there’s enough difference between my setup and yours to explain the massive differences we have in FSX performance.  

 

PMDG NGX VC View @ EDVR - 31 FPS

 

PMDG NGX Exterior View EDVR - 49 FPS

 

I am getting more than twice the FPS you received in an aircraft that is much more demanding. Something doesn't add up about the performance you are getting in FSX. I don't have FSX SE (bought it, haven't installed it yet) but I somehow doubt that I would be getting 2x the performance - that would mean 60fps in the VC and 100 fps exterior - of what I am getting now. In my opinion, your FSX install is not optimized and does not really show what FSX is capable of, even on your specific hardware.  

Don't bother discussing with OP. He clearly made up his made that FSX:SE is pure gold and everyone who's still using vanilla FSX are clearly off ground. Skepticism is not welcomed here, and if you are, then you're put on some sort of immature "ignore-list" which I remembered I used to play in kindergarden. 

 

On topic - Performance-wise, I'm not so impressed about the results, but the VAS usage (as I've said numerous times) are very interesting. I just hope we'll see the really heavy add ons such as the NGX and PMDG 777, who without doubt leaves the biggest VAS footprint. When I get a hold of this FSX version, and if VAS usage is significantly lower than before, well believe me when I say that I'm going to broadcast it to every single on here  ^_^

 

You are joking right, Jim? ...... You've raved on constantly across a number of threads since FSX:SE was released, how it's the fools choice, a waste of money why buy it, it's "just FSX", etc, etc.   

 

..... are we to believe that you're completely neutral, open-minded, and unbiased with regards to the MS FSX / FSX:SE debate?

 

...anyone buying that?   :smile:

 

 

 

Can you refer us to the thread where someone was trying to force FSX:SE on you, rather than just talking about their own experiences?

 

 

Seriously stop! Stop with that annoying attitude in which you force your modest and mediocre results on people! Accept that people naturally are skeptical and with doubt when it comes to such an big issue as this. Oh I forgot, I'm on some sort of ignore-list, am I? 

If we dont want to believe we will make ourselfes NOT believe! Some blind themselves.

The fps increase might not be mind blowing but VAS usage has definitely improved and performance feels smoother. I did my own test with a dual install and the only thing keeping me from removing the boxed fsx is asn and PMDG.

 

I run fsx in window mode and use the program called pseudo full-screen. 8n boxed fsx when I enabled this mode, sometimes it would crash or the screen would turn black before it changed. In the steam version it is instant. Only other time it changes instant with no black screen is when I use dx10.

 

Another test I used for VAS is drezwecki nycx2and NY airports. All active and starting out at klga. Is box fsx with the a2a 182 oom was instant. In steam I was flying for good 10 mins before I heard the ping. I used process explorer to monitor the VAS and it was right at 3.5 and everyone knows how demanding the drezwecki sceneries are

Been looking at VAS usage today myself.   It is indeed consistantly lower, and I've yet to experience a OOM.   (The only CTD I've ever had with FSX:SE so far was when trying out an aircraft with the Reality XP GNSs already in place).

 

I'm not using complex airliners in FSX:SE, as I mainly fly GA, so I'll not quote any VAS figures, as people just then just seem to only see all the limitations in any tests, and point out that whatever has been performed is 'invalid'.

 

But suffice to say, there does seem to be something about FSX:SE and VAS, and it's pretty exciting.

 

Please take what I've said in the spirit it's offered.  The information you and others are providing is valuable and much appreciated, and all that I'm reading taken in total does appear to be painting a very positive picture for the SE changes.  I've gathered enough information that I'd be jumping in as well were it not for some specific add-on compatibility issues which remain showstoppers, despite the fact that my FSX install has been a solid performer.

 

Scott, it's all about the spirit to me...  I did of course share my initial findings and the subsequent second test results, with the excited intention of sharing with others what I was seeing.

 

It's ok your picking up on how you perceive my reaction to some people's comments, but what about the spirit of some of those comments?  .... Some people post here as though you are their personal assistant, and your efforts just haven't been good enough..  :rolleyes:

 

What is wrong with preluding a viewpoint (especially a critical one) with something like "thanks for taking the time to do these tests. personally, I think..." (just as you did).     There are some rude people on here, with one heck of a sense of entitlement.  

 

Lastly, if you google "Realair Duke B60 2.0 frame rates, you'll see plenty of reports across the web of people struggling a bit with frame rates on this bird, due to the size of the HD textures that RealAir use on the model     - especially ATI card users.

 

The test was performed using a brand newly created FSX.CFG, and FSX_SE.CFG (with the published tweaks made to both).... so frame rate performance between both sims is validly comparable in my honest opinion.

 

The main reason I'd normally see lowish frames in the first scenario is due to what OpusFSX was doing (in both sims). Which ever way you look at it, FSX gave one frame rate results, and FSX:SE gave a considerably higher one.

 

But I'm going to back out of this thread now, as it's not a nice feeling to come to AVSIM and click on a thread I started with a sense of gloom about what's coming next! 

 

Final advice:   

 

If you fancy it; try FSX:SE

 

If you don't believe it improves upon FSX, stick with FSX.

You mention matching the published tweaks to both FSXSE and MSFSX cfgs.  I'm wondering if there are unpublished tweaks in FSXSE, that can also be made (and have been made by many) to their MSFSX cfg files and achieve similar results.  The only thing they mentioned was HighMemFix, apparently hard coded, but I'm wondering if there are more that they didn't mention.  Otherwise, what we are dealing with is a 100% improvement in your fps due to them using a new compiler alone, and that is raising some questions.  For example, I'm wondering if things like autogen density are exactly the same for a particular slider setting, or that 15% auto traffic means the same thing between both, and many other things like that.   There are lots of things in there that can be put in a .cfg file, but if they aren't there, they default to some value, and just comparing two .cfg files side by side, we wouldn't necessarily know if some underlying default value is exactly the same between the two sims. 

[TEXTUREMAXBANDWIDTH]=160  might be an example of this..

 

or is this where I have:

 

TEXTURE_BANDWIDTH_MULT=80 ?

Exactly Bert, I wasn't sure if my last post was as clear as it could have been, but that is exactly the kind of thing that I'm thinking of.  

[TEXTUREMAXBANDWIDTH]=160  might be an example of this..

 

That's what entry was in FSX.cfg when I first installed FSX-SE. Either DTG thinks it improves performance or they thought it that at worst doesn't hurt anything on most systems.

I think an interesting test would be to go apply the fixes and settings from a reputable tuning guide like NickN's, to both sims, and then see how they perform both in VAS and fps. 

I think an interesting test would be to go apply the fixes and settings from a reputable tuning guide like NickN's, to both sims, and then see how they perform both in VAS and fps. 

 

It seems to me that if FSX-SE is truely performing better than FSX-MS the aim is to figure out what needs to be done to FSX.CFG to imporove the performance of FSX-MS.  Not change FSX-SE.CFG in an attempt to lower its performance.

Scott, it's all about the spirit to me...  I did of course share my initial findings and the subsequent second test results, with the excited intention of sharing with others what I was seeing.

 

It's ok your picking up on how you perceive my reaction to some people's comments, but what about the spirit of some of those comments?  .... Some people post here as though you are their personal assistant, and your efforts just haven't been good enough..  :rolleyes:

 

What is wrong with preluding a viewpoint (especially a critical one) with something like "thanks for taking the time to do these tests. personally, I think..." (just as you did).     There are some rude people on here, with one heck of a sense of entitlement.  

 

The spirit of benchmarking is objectivity. Nobody put you to this or is asking for further work, and no one is treating you like an assistant. You took this endeavor upon yourself and posted to a public forum vital information that people new to the hobby might be inquiring about now or in the future via Google searches. Your findings then are no longer judged by your subjective feelings, they are within the scope of the community to analyze and question - without anyone being accused of rudeness and entitlement for asking those questions. It was a nice gesture to post information, but that doesn't absolve you of responsibility for its validity.  

I don't understand what what you mean by that airernie.  Why would anyone suggest tweaking FSX-SE in order to lower performance?  What I meant was if a master optimizer went to work in the cfg files of both FSX-SE and FSX-MS, trying to get the absolute best out of both of them, would we still see SE outperform MS both in fps and VAS.   The thinking is that maybe there are some unannounced tweaks in SE that either some of us already have in our MS cfgs, or that could easily be added that would put FSX-MS right up there in line with the performance that you are getting out of SE.  I think what a lot of people are saying is that some of these extremely low performance figures for FSX-MS are out of line with their experience.  They have highly tuned FSX systems, and they are plagued by neither poor fps or OOMs.  If it is true that the only tweak added by DTG was the highmemfix (which probably doesn't account for higher fps), then we are saying that simply recompiling with VS2013 is literally doubling the performance for some, and while I can't disprove that in any way, and sincerely hope it is true,  it just doesn't sound right.  Even DTG was saying that people should not expect a great difference between MS and SE.

 

@~craig~ think of it as a scientific experiment as published in a journal.  It gets subject to peer review, and that is good for everybody.  I think a lot of people are genuinely interested in your results, and the reports of others, and they just want to understand how and why some of these results are being observed and what could account for them.

I think an interesting test would be to go apply the fixes and settings from a reputable tuning guide like NickN's, to both sims, and then see how they perform both in VAS and fps. 

 

I agree. Let them both off the chain and see which one wins. I'd really like to have a sim with better performance, but before people begin mass migrations there needs to be some clarity about what the real benefits are. At this point the downsides of incompatibility are clearer than the upsides of performance. And alot of hype to confuse everything. My grandfather used to say, "don't pour out your dirty water till you find a cleaner source".