fsx-se Is it really such a leap when it comes to performance?

[pmanhart 176]

Depends on who you ask

 

SE offered me no apparent difference between FSX Deluxe

 

Same here.

 

The biggest difference for me is that some addons don't work.

 

Net loss in my opinion.

[n4gix 4,947]

Obviously if some have seen improvements that is great news but I would considering moving to FSX:SE with caution as it already has been shown that it has and will break some addons. Personally, for the hassle it just is not worth it and the following statement shows why it is not really a future proof investment of time and money.

FSX:SE does not "break some addons" at all. Those addons simply need to update their installers to accommodate potentially different Registry entries, or in some cases adapt their memory offsets if their programs depend on them.

 

This latter is nothing new of course. Every new build of FSX and P3D has resulted in Pete having to update FSUIPC to allow for new memory offsets.

[greggerm 312]

But my main question is, is it worth getting it for the improved performance and will I notice the difference when using it with Add-Ons, which can be FPS-drainer themselves ... I've heard from Simmers that you can even get double the FPS compared to the original FSX.

 

As many have said, the question you ask is quite a subjective one. It has been my experience that the visual performance differences between FSX and FSX:SE are noticeable, but are not worth some of the compatibility issues and the efforts needed to install things all over again. If you have a well-oiled and cleanly running FSX, it might not be worth the trouble right now.

 

If your FSX install is plagued by OOMs, it might be worth the while due to some of the improvements they've made, but if your install is working OK today, I'd say leave it alone for now - especially if you have to pay full price for FSX:SE.

 

-Greg

[Orbx 123]

I endorse FSX:SE fully. There is clearly a performance and VAS usage improvement, anyone trying to downplay that is wasting internet cycles, just ignore them. There were improvements made to the very old code and some glaring bugs fixed. This is not news to anybody and the evidence is tangible.

 

At their invitation, I visited Dovetail Games in Kent UK yesterday and came away impressed. Good guys there and they have a vision for flight sim.

 

FSX:SE is now installed on my dev PC and it's my go-to sim now. Enough said.

 

We are deep-testing objectflow.dll now, hopefully we'll release the FSX:SE very soon, stay tuned.

[smakar 0]

 

 

We are deep-testing objectflow.dll now, hopefully we'll release the FSX:SE very soon, stay tuned.

Great news - thank you for the hard work - I love your products!

[devgrp 1,233]

Well there you you have it guys. Another major player in the fsx community endorsing fsx steam

[torium 207]

That's interesting statements, John V. I definitely stay tuned.  :smile:

[AnkH 2,262]

Clearly, the amount of autogen in these videos are insane. Especially at the landing at the end of the second movie. Just watch. You could never do that while getting around 17-18 FPS on approach in vanilla FSX. 

 

But as far as I know, FSX:SE do not take more advantage over GPU, am I right?

 

Did you look once at the framecounter on the upper left side of the video? It has about around 12-14FPS in the approach and I am pretty sure that you also get those numbers in FSX. The discussion is about "smoothness", which you can simply not judge by a YT video. But regarding those FPS, I do certainly not see any "leap in performance", I also get around 15FPS on approach of a bigger airport. That is, what I was mentioning before the discussion switched to the observations Pete Dowson did. Most of the comments about FSX:SE being more fluent, smooth or however you call it are extremly subjective.

[devgrp 1,233]

Did you look once at the framecounter on the upper left side of the video? It has about around 12-14FPS in the approach and I am pretty sure that you also get those numbers in FSX. The discussion is about "smoothness", which you can simply not judge by a YT video. But regarding those FPS, I do certainly not see any "leap in performance", I also get around 15FPS on approach of a bigger airport. That is, what I was mentioning before the discussion switched to the observations Pete Dowson did. Most of the comments about FSX:SE being more fluent, smooth or however you call it are extremly subjective.

dude what is it with you? You can tell the smoothness from a yt video. Go watch a few videos where people have crappy computer then watch a few of froogles video and then tell me if you can't tell which is smoother. In both videos I posted you can clearly see which is smoother when I'm panning around.

 

Its like you are on a one man crusade to prove that fsx se has no improvements.

[AnkH 2,262]

You get me wrong. As I mentioned, Paul and John give hard facts by looking into the code of FSX:SE. It seems that at least some changes were made that give an explanation for the smoother appearance of FSX:SE compared to FSX:MS. All I was saying is that using a YT video is not the best argument for showing "smoothness", as a video compiled for YT (or any other video platform) basically smoothens out the "microstutters". You can never properly reproduce what you see at home on a 60Hz (or even on a 120/144Hz) monitor with a video compiled for a constant flow of 30FPS. YT video are and will never be capable of displaying microstutters lasting no longer than 1/30th of a second, it is technically impossible.

 

So, no, I do no longer ignore that FSX:SE seems to perform better, I would simply like to know WHY. But yes, I still do not like videos on YT as an argument for "smooth" or "non smooth" gameplay. Especially if they do not even show precisely the same scene. Because you insist, I stopped the video at the same moment shortly before takeoff and I marked the obvious differences on the screenshots:

 

This one is from FSX:MS, take care on the colours, the clouds, the AI traffic, the autogen etc.

 

 

Now the screenshot of roughly the same position in the FSX:SE, I did indicate the positions to look at:

 

 

Now, you can instantly see that in the FSX:SE screenshot, there is:

 

- a huge colour difference. My guess, with FSX:MS, SweetFX or something is used, which of course draws performance

- noticeably fewer AI planes are visible in the FSX:SE screenshot

- beyond the runway, FSX:MS shows noticeably more autogen buildings compared to FSX:SE

- the runway has much better display quality in FSX:MS compared to FSX:SE

- the cloud coverage is completely different in both sims

- the time of the day (position of the sun) is different.

 

Most of the above points nicely illustrate what I was always saying: it is simply not possible to reproduce two identical situation within FSX and FSX:SE. I even believe you, that you used exactly the same settings for both sims, but what the sim is displaying on the screen is different. It is ALWAYS like that. And I am sure that those little differences in your videos regarding clouds, AI traffic and autogen drawing distance can easily result in 3-6FPS difference and the difference between smooth and stuttering gameplay. I know this, because while tuning my FSX, I realized that a single AI model more or less can already result in noticable differences.

 

Verdict: to properly show the advantages of FSX:SE you should make a movie of a situation where your computer is under heavy load (such as your takeoff scenario), but you would need to pay super attention that as much as possible is really the same.

 

In the end, if you really used the same settings in above videos, my guess is that FSX:SE uses differnt settings for cloud coverage, anisotropic filtering, autogen display distance, autogen max buildings per cell and last but not least supposedly a differnt lod_radius setting (EVEN if you do not see that within the .cfg file). And those changes would already easily explain, why FSX:SE "performs better"...

[netshadoe 144]

dude what is it with you? You can tell the smoothness from a yt video. Go watch a few videos where people have crappy computer then watch a few of froogles video and then tell me if you can't tell which is smoother. In both videos I posted you can clearly see which is smoother when I'm panning around.

 

Its like you are on a one man crusade to prove that fsx se has no improvements.

 

Don't worry about the few that don't want to see.  I've switched over too and I have had great results.  I don't care so much about FPS as much as I care about the VAS shedding that the SE version does...and does well! 

 

FPS-wise, I've seen a 2-5 FPS increase at complex airports, but then again, I rarely saw below 20 FPS before, but FSX-SE in my experience is smooth, especially in the turns.

 

The big difference I've seen, and I know you've seen too devgrp, is the VAS management.  With the PMDG 777, FTX Global and Vector, REX clouds, and early ASN beta tests, I got my VAS at my starting airport up to 3.9+GB - I had about 150MB left - using all the tricks I could think of to load it up...I was also using 100% AI with about 250 traffic files, a mix of WOAI and MilitaryAI. 

 

With this 3.9GB "load" with the PMDG 777 and ASN, I was able to taxi out and takeoff.  30 minutes after departure, my VAS was down to 3.5GB.  By the end of the first hour, I was teetering at about 3.1-3.2GB.

 

Six hours later (I don't get the terrain.dll and G3D.dll crashes for some reason) I landed at LAX and taxied in and parked at the gate...VAS was only at 3.6GB.  So, I lost 300MB net over the flight.  In FSX, I would have OOMed at departure because of the VAS load.

 

I'm happy with my switch.  If others don't want to see, don't bother wasting your breath.  It's like they're threatened...it doesn't make sense to me either, but I digress.  Those of us with open minds will listen, evaluate and decide to choose for themselves instead of blasting those of us who are basically the guinea pig for this new iteration of this sim.

[AnkH 2,262]

Well, things getting clearer now, I continued to watch your movies and sorry, IF you really used the same settings, then FSX:SE plays some easy tricks. I now rather think that you played tricks with us, simply by changing the settings within FSX:SE to illustrate what you WANT to see.

 

FSX:MS. Take care of the amount of autogen buildings, especially where the big warehouse zone ends and the small houses start:

 

 

Now compare the amount of autogen buildings in above picture with the screenshot below, taken from the FSX:SE movie:

 

 

You must be kidding if you say that you did not realize this. FSX:MS shows about ten times more autogen buildings compared to FSX:SE. No wonder, it seems to be "smoother".

 

I'm happy with my switch.  If others don't want to see, don't bother wasting your breath.  It's like they're threatened...it doesn't make sense to me either, but I digress.  Those of us with open minds will listen, evaluate and decide to choose for themselves instead of blasting those of us who are basically the guinea pig for this new iteration of this sim.

 

This is not the issue. If above videos show really the same settings, FSX:SE simply BUYS more FPS by reducing autogen, cloud coverage and stuff and you guys do not even notice it. I hope, that at least Paul and John did realize this and took care about those facts, then there is still some hope, that FSX:SE really performs better. But with a ridiculous comparison of two videos obviously showing huge differences in autogen settings, cloud coverage and AI traffic, nothing can be proven, sorry to say...

[AnkH 2,262]

More evidence found that those two videos are a simple show off. You can download both .cfg files from the YT links, I did compare them. Differences are:

 

1. The FSX.cfg shows MultiSamplesPerPixel=8 and MultiSampleQuality=8 while those two entries are missing in the FSX:SE.cfg. Surprise, Surprise, this is a huge load on the FSX compared to the FSX:SE

 

2. Autogen-Density entry is the same. This somehow let me think that Dovetail "hardcoded" the reduced amount of autogen somewhere else (similar to HIGHMEMFIX, which is not needed in the FSX:SE.cfg anymore).

 

3. TERRAIN_MAX_AUTOGEN_BUILDINGS_PER_CELL=1000 and TERRAIN_MAX_AUTOGEN_TREES_PER_CELL=1500 entries are missing in the FSX:SE.cfg.

 

4. WATER_EFFECTS=7 in the FSX.cfg whereas in the FSX:SE.cfg, there is "only" WATER_EFFECTS=6, this also reduces the load on the system

 

5. TEXTURE_BANDWIDTH_MULT=40 in the FSX.cfg, TEXTURE_BANDWIDTH_MULT=160 in the FSX:SE.cfg. No comment...

 

6. CLOUD_COVERAGE_DENSITY and CLOUD_DRAW_DISTANCE are identical. As you can see a clear difference in the videos, I fear that Dovetail also "hardcoded" this setting outside the FSX:SE.cfg...

 

7. FIBER_FRAME_TIME_FRACTION=0.40 is found in the FSX.cfg, this entry is absent from the FSX:SE.cfg. As a matter of fact, the lower this number, the more FPS you get. Default value (and therefore what FSX:SE uses) is 0.33...

 

So, now tell me again how you should be able to properly compare those two sims using the posted videos. There are to many differences obvious to see from the video itself and from the .cfg files. No, this does not convince me at all and especially the fact that there is so much less autogen and cloud coverage even if the settings are identical within the two configs, let me raise an eyebrow on what the heck Dovetail Games did to their FSX:SE variant...
 

[AviatorMoser 353]

Oh geez, the cherry picking begins.

[netshadoe 144]

Goodness...glad you are taking so much time proving your....I don't really know what you're trying to prove, really. 

 

If you like your FSX, then keep it.  No one is ripping out your fingernails to switch over.  That's why they make different makes of cars and different colors...everyone is different in what they're looking for.  I drive a VW Jetta Diesel...you going to tell me that I would get 0.5% better mileage if I would have gotten a Diesel Golf?!?!  It's that nitpicky??

 

Instead if wasting your time proving...again, I don't know what you're trying to prove...whatever...spend your time flying your sim you're happy with, and the rest of us will fly what we want, DoveTail or not.  It's about enjoying the sim...not counting how many autogen trees I have in my frame...I'm not that OCD.

Oh geez, the cherry picking begins.

 

Uh huh!!    :Rolling Eyes: