Based on the pics,   yes there are certainly less vehicle's in the Steam Edition 

 

 

I don't know what style you are referring to, must be some U.S. thing. Pete is right, this chap has been trying to find fault with Steam since it came out. I wish he would shut up.

 

Ah not at all,  Jim tells it how it is.

 

 

When hes not  FSX, 

xp10.jpgh by *poppet*, on Flickr

 

I heard hes an X-Plane fan,  you should ask him,  from a safe distance though   :Big Grin:

We are nitpicking here. The difference in density of default vehicles is negligible. I bet that if you use any traffic addon we won't be able to tell them apart. The fps difference however is another story.

Well, it looks like I came closer to the truth: FSX-SE seems to have a lower rendering radius for ground traffic!  If I move forward in slew mode, the ground cars simply pop up at a short distance, but beyond that distance, there is almost no ground traffic at all. FSX-MS, on the other hand, displays all ground traffic at a much larger radius, hence the lower FPS.

Insert in Scenery section:

 

[sCENERY]
SmallPartRejectRadius=0

 

and suddenly you can see all. evenboats.

 

cheers!

Well, it looks like I came closer to the truth: FSX-SE seems to have a lower rendering radius for ground traffic!  If I move forward in slew mode, the ground cars simply pop up at a short distance, but beyond that distance, there is almost no ground traffic at all. FSX-MS, on the other hand, displays all ground traffic at a much larger radius, hence the lower FPS.

 

Question: are you using the same CFG file in both FSX-MS and FSX-SE, as I am? If not, that's what you need to do before doing these comparisons.  Also check add-ons which have added or altered ground traffic like AES and GSX.

 

There is really no way DTG have turned anything down at all.Why should they? To earn more £4.99's in the sales?

 

You are not doing the comparison on an equal footing.

 

There is no reason for them to turn anything down. The recompiled code is much more efficient on today's processors, compared with the old compiled code made over 9 years ago using a compiler which was over a year older at the time! The code optimisations to suit threading and parallel pipes has come on a long way, taking better advantage of today's processor architectures.

 

I can now run at dense airports with all scenery sliders at max (excepting Autogen which I still keep one notch down), and MyTrafficX airline traffic at 100% -- I could not use UK2000 or Aerosoft Heathrow at any higher trsfic settings than 40% and 55% respectively in FSX-MS, and both versions of Heathrow gave jerky performance in any case, not so in FSX-SE.

 

Why don't those who doubt all this not even try it? Vested interests? In what, exactly? Why not accept that is performs better all round, both frame rate-wise and smoothness. There are enough FSX-SE users now who say so, and the doubters are mostly those who've not even tried it (and one prominent one in particular).

 

And DTG's programmers are working hard on it too. The current private Beta, for developers to get ready for, is the 7th version since the release. Each has had both some newly introduced bugs fixed AND, more importantly, a number of original FSX-MS bugs, like those in G3D (of which FSUIPC used to have to shield users from but one) and in Terrain, some of which were, admittedly, caused by slightly over-zealous attempts to free up VAS too early).

 

Personally I feel DTG have given FSX a real new lease of life and I am over the moon with how my cockpit is now performing.

 

Pete

Apparently the same topic was discussed few days ago on the Steam forum and the problem was solved with the same solution suggested by abranpuko:
http://steamcommunity.com/app/314160/discussions/0/618457398971552821/

Apparently the same topic was discussed few days ago on the Steam forum and the problem was solved with the same solution suggested by abranpuko:

http://steamcommunity.com/app/314160/discussions/0/618457398971552821/

 

Thanks. That helps seeing AI in the air, for sure. Now I can see them a lot further away than I could in FSX-MS! So now I need to see if there's any impact at Heathrow (my test location)!

 

Pete

If you disable the ground traffic in both sims, the performance should match if your hypothesis (re: ground traffic fps impact) is true. Can you test that case? I lack the "SE" part for that.

 

Besides, it's always encouraging to see folks testing and questioning certain aspects of, at times, "given" knowledge. Doesn't have to be successful all the time. It's the thought that counts. :smile:

Our Jim does it with a certain sassy style

 

 

Well, it looks like adding the SmallPartRejectRadius entry brings back the vehicles, but gone are the FPS gains.

 

 

It is good to know that at least the vehicles are not completely gone in FSX-SE, but only obscured beyond a small radius. This is something I even appreciate, since I don't need to see service vehicles from a far away distance during taxi or landing if I can trade that for more FPS and/or less stress on the CPU. I felt kind of... cheated... by FSX-SE when I first saw the FPS gains at seemingly no difference in visual quality. Only further investigation revealed that there was no magic at play, after all.

 

By the way, I have played around with the fsx.cfg of the DVD version and found that adding SmallPartRejectRadius=3 or 4 also hides the vehicles in the distance (and consequently increases FPS to levels comparable with what I previously saw in FSX-SE). Having the opportunity to modify the distance per the entry in the DVD version, the improvement in FPS as a result of hidden vehicles farther away is nothing exclusive to the Steam Edition.

Alright, I did some more testing. I installed MyTrafficX 5.4b on both sims and set the Airline and the GA Traffic slider to 100%. You can expect that these traffic settings put an enormous stress on the CPU. I used the "simple model" of the AI planes via MyTraffic Communicator and disabled the jetways (otherwise it would be hell for FPS). I set the ground traffic density slider to zero in both sims to avoid the FPS impact by the different rendering radius of vehicles. I also removed the SmallPartRejectRadius entry in both cfg's.

 

Then I opened the save file that I created in KATL and waited 5 minutes with each sim for airline traffic to set in motion and for FSUIPC to record the average frame rate and other parameters before exiting the sim exactly on time. With all this, the picture in Atlanta looks like this after a while in both versions:

 

 

As you can see, ALL parking spots are populated with airplanes, and there are many more invisible airplanes that are en route up in the sky (or at nearby airports).

 

The FSUIPC log shows the following:

 

FSX-MS:

 

********* FSUIPC4, Version 4.939j by Pete Dowson *********
Running inside FSX on Windows 7
Module base=62FA0000
User Name=""
User Addr=""
FSUIPC4 not user registered
WIDEFS7 not user registered, or expired
       78 System time = 04/04/2015 01:25:24
       78 FLT path = "C:\Users\[username]\Documents\Flight Simulator X Files\"
      561 Trying to connect to SimConnect Acc/SP2 Oct07 ...
      561 FS path = "D:\Microsoft Games\Microsoft Flight Simulator X\"
      717 LogOptions=00000000 00000001
      717 --- CONTROLS timer memory location obtained ok
      717 --- SIM1 Frictions access gained
      717 --- FS Controls Table located ok
      717 --- Installed Mouse Macro hooks ok.
      717 --- Wind smoothing fix is fully installed
      717 --- G3D.DLL fix attempt installed ok
      717 SimConnect_Open succeeded: waiting to check version okay
      717 Trying to use SimConnect Acc/SP2 Oct07
     2262 Running in "Microsoft Flight Simulator X", Version: 10.0.61637.0 (SimConnect: 10.0.61259.0)
     2262 Initialising SimConnect data requests now
     2262 FSUIPC Menu entry added
     2293 D:\Microsoft Games\Microsoft Flight Simulator X\FLIGHTS\OTHER\FLTSIM.FLT
     2293 D:\Microsoft Games\Microsoft Flight Simulator X\SimObjects\Airplanes\Aircreation_582SL\Aircreation_582SL.AIR
    11700 Weather Mode now = Theme
    11700 C:\Users\[username]\AppData\Roaming\Microsoft\FSX\Previous flight.FLT
    20233 c:\users\[username]\documents\flight simulator x files\KATL.FLT
    31169 System time = 04/04/2015 01:25:55, Simulator time = 12:43:27 (17:43Z)
    55037 Starting everything now ...
    56254 Advanced Weather Interface Enabled
   350019 Sim stopped: average frame rate for last 300 secs = 31.3 fps
   355822 System time = 04/04/2015 01:31:20, Simulator time = 12:48:26 (17:48Z)
   355822 *** FSUIPC log file being closed
Minimum frame rate was 24.0 fps, Maximum was 36.0 fps
Minimum available memory recorded was 1560Mb
Average frame rate for running time of 300 secs = 31.3 fps
G3D fix: Passes 18682, Null pointers 0, Bad pointers 0, Separate instances 0
Memory managed: 133 Allocs, 133 Freed
********* FSUIPC Log file closed ***********

 

FSX-SE:

 

********* FSUIPC4, Version 4.939j by Pete Dowson *********
Running inside FSX Steam Edition on Windows 7
Module base=5F820000
User Name=""
User Addr=""
FSUIPC4 not user registered
WIDEFS7 not user registered, or expired
       31 System time = 04/04/2015 01:06:17
       31 FLT path = "C:\Users\[username]\Documents\Flight Simulator X Files\"
       47 Trying to connect to SimConnect Steam ...
       63 FS path = "C:\Program Files\Steam\steamapps\common\FSX\"
      187 LogOptions=00000000 00000001
      187 --- CONTROLS timer memory location obtained ok
      187 --- SIM1 Frictions access gained
      187 --- FS Controls Table located ok
      187 --- Installed Mouse Macro hooks ok.
      187 --- Wind smoothing fix is fully installed
      187 --- G3D.DLL fix attempt installed ok
      187 --- TERRAIN.DLL fix attempt installed ok
      187 SimConnect_Open succeeded: waiting to check version okay
      187 Trying to use SimConnect Steam
      952 Running in "Microsoft Flight Simulator X", Version: 10.0.62608.0 (SimConnect: 10.0.62608.0)
      952 Initialising SimConnect data requests now
      952 FSUIPC Menu entry added
      983 FLIGHTS\OTHER\FLTSIM.FLT
      983 simobjects\airplanes\Aircreation_582SL\Aircreation_582SL.AIR
    30202 Weather Mode now = Theme
    30202 c:\users\[username]\documents\flight simulator x files\KATL.FLT
    41949 System time = 04/04/2015 01:06:58, Simulator time = 12:43:27 (17:43Z)
    76862 Starting everything now ...
    78110 Advanced Weather Interface Enabled
   371953 Sim stopped: average frame rate for last 300 secs = 32.0 fps
   377335 System time = 04/04/2015 01:12:34, Simulator time = 12:48:26 (17:48Z)
   377335 *** FSUIPC log file being closed
Minimum frame rate was 26.3 fps, Maximum was 35.8 fps
Minimum available memory recorded was 1608Mb
Average frame rate for running time of 300 secs = 32.0 fps
G3D fix: Passes 18620, Null pointers 0, Bad pointers 0, Separate instances 0
TERRAIN fix: Passes 1013, Null pointers 0, Bad pointers 0, Separate instances 0
Memory managed: 135 Allocs, 135 Freed
********* FSUIPC Log file closed ***********

 

As you can see, FSX-MS average FPS is 31.3, and FSX-SE is 32.0. FSX-SE is a little bit better, but based on these numbers, there is no way for me to conclude that the Steam Edition runs "much better" than the DVD edition with or without all the traffic. Again, no add-ons are installed except MyTraffic and FSUIPC. I kept separete cfg files for both sims, because I thought that this was the best way avoiding each sim overwriting the file on their own. I thought that setting the in-game settings at identical positions (except the global resolution slider, which is extended to 4096 in FSX-SE) would be sufficient for a fair comparison (in addition to adding the HIGHMEMFIX entry in FSX-MS, which I admit I haven't done).

 

To me, there might have been incentives to stick with the SE version if all my add-ons would run on it, but Concorde-X is one for example that doesn't work with SE and won't for a while according to the developers. I don't harbor resentment against DTG's commitment to further improve FSX-SE or against users who have switched to that edition completely, but I would caution against overhyping the SE edition vs. the DVD version as so much better in terms of FPS and VAS. Like I figured out today, DTG has applied some unannounced "tweaks" (SmallPartRejectRadius) that boost the FPS at the cost of shorter display radius of vehicles. Earlier, I thought that FSX-SE's performance was naturally better, but it looks like with manipulating the aforementioned radius entry, FSX-MS's performance and display can be brought on par with that state.

Bonjour,

 

DTG never claimed that FSX SE has better performance.

You can read this in this post  Existing FSX Owners, check here first [Official

http://steamcommunity.com/app/314160/discussions/1/626329820730080279/

 

 

This is a re-release of FSX Gold Edition so you should not expect dramatic differences.    

However, we do appreciate the current benefits for existing users are modest.  

 

 

With FSX , I made a lot of tweak, but with heavy load (PMDG, scenery...) I always have the feeling that there were bottleneck in the process even with sliders not too high . Now with FSXSE, slider high, except autogen, the sim is like some people say "smooth like butter (fly)" :smile: 

 

I do not know the reason, but I think, as Pete said, that the recompilation, with new tools, makes a better use of the thread and core. That can explain the disappearance of bottleneck and the feeling of smooth.

 

Best regards

Claude

 

 

 

As you can see, FSX-MS average FPS is 31.3, and FSX-SE is 32.0. FSX-SE is a little bit better, but based on these numbers, there is no way for me to conclude that the Steam Edition runs "much better" than the DVD edition with or without all the traffic. Again, no add-ons are installed except MyTraffic and FSUIPC.

 

I also get similar performance when the Sim is not heavily loaded, as it evidently is not in your two tests. Your average fps of 30+ shows this.

 

In my heavily loaded tests (very heavy airport, lots of AI traffic, GSX vehicles, ASM complex weather) I would only get frame rates between 15-25 in FSX, deteriorating over time to 9-15 as the AI built up taxiing, taking off and landing.

 

With FSX-SE, in exactly the same situation, with the same CFG settings (and, yes, I have always used the SmallPartRejectRadius option, set to 3 I think (but I'd need to check that -- not tried 0 before, but I will be doing that), I get 30-40 deteriorating to 19-30 eventually. 

 

More to the point, taxiing along Heathrow's runways is very much smoother, immediately noticeably so just by looking at the sides of the taxiways where, in FSX-MS, you can see the jumps in the scenery drawing. In turns FSX-MS is awfully jerky, but smooth in FSX-SE.

 

I would agree that if you never stress your FSX-MS to the point where it is jerky, where the frame rates are so poor, then there maybe little point in changing to fSX-SE. Except possibly for the changes they made, such as fixing more G3D.DLL crashes than FSUIPC was able to trap, and freeing up VAS more consistently.

 

On the crashes front, with the version of FSX-SE soon to be released on Open Beta, FSUIPC does not patch either G3D or TERRAIN DLLs because the bugs it was trying to trap are fixed.

 

ON top of this, DTG are being very responsive to bug reports, fixing them as they are nailed down. To help in that process, if FSX-SE does crash there's a minidump file automatically produced (.MDMP) in the FSX folder, which helps DTG track it down. You have to send it to their support email address with details of the circumstances and add-ons in use.

 

I see the only disadvantage of using FSX-SE is that some add-ons, those which hook into the innards of FSX, like FSUIPC, ASN, GSX, AES and some add-on aircraft gauges, may need changing for each FSX-SE update, and there may sometimes be delays in this. But this was true also of the MS updates SP!, SP2, Acceleration -- it was just that those were months apart, whilst DTG are reacting much quicker (7 releases, 4 of them private Betas only,in four months!).

 

I should add that the changes needed, update to update, are generally minor and easy to allow for. With FSUIPC, despite using six assembly code hooks and 20-odd different function calls into FSX, it is done within a couple of hours of receiving a Beta release.

 

Pete

Well, it looks like adding the SmallPartRejectRadius entry brings back the vehicles, but gone are the FPS gains.

 

[...]

 

It is good to know that at least the vehicles are not completely gone in FSX-SE, but only obscured beyond a small radius. This is something I even appreciate, since I don't need to see service vehicles from a far away distance during taxi or landing if I can trade that for more FPS and/or less stress on the CPU. I felt kind of... cheated... by FSX-SE when I first saw the FPS gains at seemingly no difference in visual quality. Only further investigation revealed that there was no magic at play, after all.

 

By the way, I have played around with the fsx.cfg of the DVD version and found that adding SmallPartRejectRadius=3 or 4 also hides the vehicles in the distance (and consequently increases FPS to levels comparable with what I previously saw in FSX-SE). Having the opportunity to modify the distance per the entry in the DVD version, the improvement in FPS as a result of hidden vehicles farther away is nothing exclusive to the Steam Edition.

 

Default value is 1.

Some people recommend to set it to '2', '3' or even '4' to have FSX not drawing autogen objects smaller than 'X' pixels in size. A higher value improves FPS at the price of a lesser AI aircraft and object visibility and an increase of popping.

 

I always had this value set on '1' both in FSX-SE and in my previous FSX classic. Last night I tested an approach on Sunskyjet KPHL (which I consider heavy on frames) with the value set to '0'.

 

Good news: impressive resolution, all aircraft and objects visibile, negligible or absent popping.

Bad news: significant drop in FPS.

As usual, here we have a question of balance.

 

Note: all sliders maxed out except autogen density, traffic set to 30%

 

 

Doing some SmallPartRejectRadius testing here.

 

- It does not affect autogen or highway traffic, it only affects AI models (aircraft, airport vehicles, boats)

- The default value must be 4 or so => Set it to higher than 4 if you need a performance boost at airports

 

 

I've set mine to "6" for now for a 0.5 FPS gain over NYC.