Hi,

 

Excellent report - Forgetting FPS, could you say which one actually ran the smoothest? i.e. least amount of stuttering / tearing if any!

 

Regards,

 

 

Ted.

 

Its hard to say, to be honest. I think that with this much autogen overkill both struggled equally as we all know the code has a performance ceiling of sorts. I personally will probably be flying with cars off to make things a little smoother and will probably be trying DX10.

 

However, I dont know if it was when FSX: SE "released" its texture cache, when I turned on the PDZ approach just downwind from KONT, the sim smoothed out significantly. I dont know the black magic behind it but it was a butter smooth approach. 

 

Thanks for the report.

 

For the fps, did you try running your benchmark in full screen?

 

The FSX-SE picture on the runway from the cockpit might have lower fps because your point of view was a little lower and in FSX-SE you also had a part of the cockpit to draw on screen.

 

Yes, the first 2 sets of shots you see were in full screen. I only windowed for the Memory shots. The FPS was pretty much the same 95% of the time to be honest. A little blade of grass or a couple of knobs didnt ruin the data for me. 

 

The more the merrier -when will you publish your "far more compreshensive and empirical tests" in detail?

 

I also look forward to that report! 

 

I have to "respectively" disagree with this statement. Although there may be some who do this, one of the main purposes of forums is to share ideas and experiences, not just ask for help when having issues. It's the same concept with reading reviews of movies or restaurants - yes, individual tastes differ, but I can get a general idea from other experiences.

 

I too have purchased but not installed FSX-SE. But given that I've recently encountered an issue preventing FS2CREW audios from working (used for PMDG aircraft), and the recommended fix by the vendor is to reinstall W7, I am planning on doing a fresh reinstall this weekend, but then installing FSX-SE. Yes, I know PMDG aircraft will not work (yet), but I have plenty of other aircraft that do. And given that I often encounter OOMs (especially with MegaSceneryEarth and add-on airports), FSX-SE gives me hope, thanks to this post. And if all else fails, I can always go back to FSX from my DVD - only a couple of days lost.

 

Happy Holidays to all!

 

Happy Holidays to you as well! 

 

Idk what's going on with Avsim. A fellow flight simer took time out of his very busy schedule just to have people knock his benchmark comparison? What a bunch of ungrateful people. People here walk around with this "chip" on their shoulders. Get over yourselves and get off your high horses! Lets help each other.

 

 

Thanks for the support, to be honest I knew there would be some cat fights and at the end of the day we are still a community all just waiting for our 64bit. 

Provided that someone didn't already mention it:

 

You can monitor remaining VAS with FSUIPC (any version).

http://forum.avsim.net/topic/431663-vas-fsuipc-and-fsx/

 

Might warrant another test run as most users here seem to be most interested in VAS.

 

 

With the exact same settings, FPS in FSXSE seemed to be lower for me than in legacy when looking at performance intensive scenes (busy terminals from far away) and slightly higher when not.

In my JFK test scene, I got ~12 FPS with the variance between FSXA and FSXSE being about 0.5 FPS. After kicking out the internal frame limiter and using the NVidia driver to limit FPS, I got a whopping 5-6 FPS increase in FSXSE in the same scene (at the cost of higher fluctuations).

Can you test it in DX10 preview? I was also having FPS issues despite having a $2000 computer but last week I tried using DirectX10 preview and my FPS went up by at least 10 in FSXA. I wonder how well it would do in FSXSE. 

 

Thanks for the benchmark,

Regards,

Provided that someone didn't already mention it:

 

You can monitor remaining VAS with FSUIPC (any version).

http://forum.avsim.net/topic/431663-vas-fsuipc-and-fsx/

 

Might warrant another test run as most users here seem to be most interested in VAS.

 

 

With the exact same settings, FPS in FSXSE seemed to be lower for me than in legacy when looking at performance intensive scenes (busy terminals from far away) and slightly higher when not.

In my JFK test scene, I got ~12 FPS with the variance between FSXA and FSXSE being about 0.5 FPS. After kicking out the internal frame limiter and using the NVidia driver to limit FPS, I got a whopping 5-6 FPS increase in FSXSE in the same scene (at the cost of higher fluctuations).

 

I agree with your findings and my tests show the exact same thing. The memory seems to be slightly higher in lower demand situations and is managed dramatically better in high demand scenarios. 

 

I personally have always used Private Working Set memory found in the Task Manager as my way of monitoring the limits of 4GB of addressable memory in a 32bit application and as my indication of an OOM. As MGH previously referenced in this thread, the working set of a program is a collection of those pages in its virtual address space that have been recently referenced. It includes both shared and private data. The shared data includes pages that contain all instructions your application executes, including those in your DLLs and the system DLLs. As the working set size increases, memory demand increases.

 

 

 

Can you test it in DX10 preview? I was also having FPS issues despite having a $2000 computer but last week I tried using DirectX10 preview and my FPS went up by at least 10 in FSXA. I wonder how well it would do in FSXSE. 

 

Thanks for the benchmark,

Regards,

 

I will very soon. 

 

 

Anyone who posts here with a title "
The Conclusive Performance Benchmark"
Is either very brave or very new to Avsim...or both....welcome to the never ending world of "no good deed goes unpunished", other wise known as the internet experts forum :-)

 

Ain't that the truth !!!  I am so tempted to start a new thread titled "An Irrefutable and Indisputable Interpretation of Flight Simulation EULAs" before leaving town for Christmas.

 

Kudos to the O.P.  Maybe not "conclusive"... but very good information nonetheless.

 

M.D.

The problem is Steam version had a couple of fixes added by default, ie

 

* Increased TEXTURE_MAX_LOAD to 4096 up from 1024
* Changed HIGHMEMFIX=1

* UIAutomationCore.dll work-a-round is now supplied by default with FSX:SE. This fixes a number of crashes relating to UI interaction on 64-bit versions of Windows

 

If these where not set in FSX than it makes the test invalid.

The problem is Steam version had a couple of fixes added by default, ie

 

* Increased TEXTURE_MAX_LOAD to 4096 up from 1024

* Changed HIGHMEMFIX=1

* UIAutomationCore.dll work-a-round is now supplied by default with FSX:SE. This fixes a number of crashes relating to UI interaction on 64-bit versions of Windows

 

If these where not set in FSX than it makes the test invalid.

 

So you clearly didn't read the original post...

Anyone who posts here with a title "

The Conclusive Performance Benchmark"

Is either very brave or very new to Avsim...or both....welcome to the never ending world of "no good deed goes unpunished", other wise known as the internet experts forum :-)

 

 

Ain't that the truth !!!  I am so tempted to start a new thread titled "An Irrefutable and Indisputable Interpretation of Flight Simulation EULAs" before leaving town for Christmas.

 

Kudos to the O.P.  Maybe not "conclusive"... but very good information nonetheless.

 

M.D.

 

LOL... I have been :ph34r: creeping around these forums for many many years before they required you to have an account to visit. So I knew what I was getting into.  :He He:

I believe an outside view does not render any more than if you are inside the cockpit. It renders less, because you are not rendering the virtual cockpit textures. Inside the cockpit it doesn't matter if you have a limited view, the terrain radius is still rendering everything on the ground, and adding the vc textures ontop of that. Looking forward to new benchmarks, thank you for doing this.

Can you test it in DX10 preview? I was also having FPS issues despite having a $2000 computer but last week I tried using DirectX10 preview and my FPS went up by at least 10 in FSXA. I wonder how well it would do in FSXSE. 

 

Thanks for the benchmark,

Regards,

 

So I dont have the time to go through all the image uploading and stuff. But I did want to report back that I tried DX10.

 

Overall, the FPS didnt really change at all on the ground, but in the air and once the autogen really started going, FPS increased 7-10 frames. Once again, I think this is a performance ceiling thats deep coded in FSX.

 

Although frames performed well, Memory usage was down DRAMATICALLY throughout the sim about 350-500 MB's.

 

Over KLGB for example 2.34 GB vs the original 2.79 GB and 12 NM from PDZ it was 2.23 GB vs the original 2.64 GB and finally approach was 2.8 GB vs 3.1 GB originally. So a solid savings in Memory usage with DX10

So I dont have the time to go through all the image uploading and stuff. But I did want to report back that I tried DX10.

 

Overall, the FPS didnt really change at all on the ground, but in the air and once the autogen really started going, FPS increased 7-10 frames. Once again, I think this is a performance ceiling thats deep coded in FSX.

 

Although frames performed well, Memory usage was down DRAMATICALLY throughout the sim about 350-500 MB's.

 

Over KLGB for example 2.34 GB vs the original 2.79 GB and 12 NM from PDZ it was 2.23 GB vs the original 2.64 GB and finally approach was 2.8 GB vs 3.1 GB originally. So a solid savings in Memory usage with DX10

Oh, I see. I may have to do more testing myself. I will also try it with the directX10 fix available on avsim.

Thanks

 

* Changed HIGHMEMFIX=1

 

 

I believe this one is hard-coded, no need to add.

(Pls correct me if I'm wrong)

 

Dirk.

I believe this one is hard-coded, no need to add.

(Pls correct me if I'm wrong)

 

Dirk.

In FSX SE it is. But in vanilla MS FSX it needs to be added. Most addons can't work in MS FSX without the fix.

Also, thanks for the report, and ignore the whiners and whingers

 

It really amazes me that some people's intent on this Forum, Bofajking was simply making a comparison between FSX and FSX SE, It most likely took him a lot of his personal time to compile his post, then for people to come on and slag him off for it.

 

I haven't commented on this thread so far and I had no intention either, but do I feel the need to say that we shouldn't call the people who have a criticial note "whiners", "whingers" or ones that "slag him off". Just because someone puts a lot of effort into something, which all us of appreciate, shouldn't stop others from  sharing their opinions. Their opinions are just as imporant as this test.

 

Just my 2 cents.

I look at the emergence of the Steam version as being a very good thing. 98% of the people who buy it won't get into it.. But I'm sure that 2% who do, will sink as heavily as we have, and that 2% on 'steam' could be a huge number.. This benefits the developers because of more customers.. It benefits us, because the hobby becomes less niche.. I mean there are tons of people that buy stuff during Steam sales, and they never even use the product.. the low prices, just make them jump at the sale.. What better way to be selling your product, than to people who don't even really want it!