Thanks for all you do for us serious FSX glassy performance seekers!!Dave

Hi Dave,It's a known fact that the internal FPS limiter will produce a much smoother experience, this is because when you LIMIT your frames INTERNALLY FSX doesn't use the other threads to load terrain, it happens in the MAIN scheduler, so there is no threads/fiber syncronization going on.When frames are set to UNLIMITED FSX offloads this from the main scheduler, you DO GET increased performance but it is not 100% stable, you get some sligh stuttering. A way to counteract this stuttering is to reduce (or remove) FIBER_FRAME_TIME_FRACTION if present.You can test this by yourself by running LIMITED frames inside FSX and setting FIBER_FRAME_TIME_FRACTION to 0, you get the silky smooth experience, stable FPS, no impact on FPS and lots of blurries. (see Word Not Allowed's FFTF=0 thread)

Don't worry, the settings set by the tool are fine. There is a very slight performance increase when you enable the AC_SELF_SHADOW option and leave AIRCRAFT_SHADOWS disable (0) the gain is marginal. about 2-3% so not really an issue, you'll probably won't even notice any gain in FPS.I have a baseline test that use to determine what settings affect performance. This test is not based on observation, but measurement using external tools while looping a situation (replay) over 4 times. I always discard the 'first' loop, and measure the 2nd, 3rd and 4th. If they ALL match and show any diff I then consider a setting to have an impact.I use FS Recorder (single frame record) and PerfMon.exe using a laptop over the network to make this measurements. The AC_SELF_SHADOW is one of those mistery settings that provide a slight gain whitout a clear explanation of why it is so.

Are these test done from the cockpit?

Good question from freebird77 there, I double it. Strange thing, hearing about aircraft self shadows actually enhancing performance (don't worry, "strange" doesn't mean that I don't trust you, it's just .. strange). Sad to hear that the aircraft shadows have to be off then, the offer quite some atmosphere in my eyes. But hearing about marginal savings there, I'll stay with the atmosphere instead of some 2-3%. But surely an interesting finding from you.*******, I did some testing on this transparency AA thing again. The 10ft high stutters occurred with different planes so far but I think that they are scenery related. I e. g. hat die CS 707 to a standard airport in FSX and got the stutters. Later, I took the same bird from that standard airport to Orbx Melbourne 2.0 and landed as fluently as ever, no stutters at all.So I don't think that your Level D 767 has anything to do with it but that the actual scenery does. I therefor can't report about a general transparency AA on = 10ft high stutters relationship. Seems more like it's there on some sceneries while others never struggle.Also, my cure for this wasn't a real one as either "enhance" and "override" in the AA policy of the Nvidia driver doesn't change this behavior so far. Another question would be, how AMD cards behave on this.I hope you had some nice Christmas time there. Seems like you never pause investigating. :(

*******, The real problem is how to get the best, smoothest experience when you have to run in windowed mode. In windowed mode, unless you really have a screaming machine and don't want much eye candy, you can't always get more than 60 fps and therefore, you must live with 30 (60/2) fps, typically. Unfortunately, while that appears to be fine for an aircraft that moves straight forward most of the time (I hope!), doesn't fly all that low (most of the time), and doesn't spin around on it's yaw axis (most of the time). Obviously, I am describing the usual arena for a helicopter which is my target flight sim environment. With a helicopter, 30 fps shows quite a bit of two frame "stutter" (not like FSX stutters) - it just means that one frame I was at 22.5 degrees in heading and the very next frame I was at 22.8 degrees or so and my eye sees that very definitely. I know that I have plenty of performance and things are tuned/tweaked very well from many of your recommendations because if I fly in FullScreen, I am nearly always at 60fps and the image is glassy. Also, if I temporarily set ForceWindowedVsync=0 for a test run, the fps is 50->60 in windowed mode. But when I run in windowed, Vsync-on mode, the best I can run at is 30 fps an even though the variance printout shows less than 1 % (using either external or internal frame limiting), you get the stutter in fast yaw motion especially. That is why I discussed running with a faster monitor update rate so maybe 37.5fps looks better. Obviously, it would be great to limit at 60fps in windowed mode, but I don't have that kind of performance. It's a problem - without Vsync the image looks awful, but with it, the performance falls from nearly 60 to 30 and the next frame stutter is quite visible. Well, enough chatter, thanks for all your help. I really appreciate the extra performance due to your research that many of us are seeing. The interesting part is that I see much better performance with my own oclocked Q6600/NV8800GT as well as trying to improve a 980x/470 system driving three monitors. We are soooo close to full fidelity flight sim performance (I helped build them for Boeing Simulator Engineering in Seattle where we spent millions on our cockpits and used a PC per 45 degree channel x 5). You have really helped!!Dave

Are these test done from the cockpit?

From the VC, yes. I'll be happy to have someone explain to me 'why' this has any impact on perf, it IS marginal, yes... but why would it be any impact?

From the VC, yes. I'll be happy to have someone explain to me 'why' this has any impact on perf, it IS marginal, yes... but why would it be any impact?

That kind of behavior seems odd.In DX9 the AC_self shadows take there toll on frames (if any) when they get rendered, in -->outside view (I run with them enabled). As for AC shadows, those are for rendering the aircrafts Sun perspective silhouette shadow over the terrain, also usually not seen unless in outside view.

*******, OK - now we are really getting somewhere. In windowed mode, with FFTF=0.0 and Vsync-on - no limiting or external limiter, I get a fairly smooth appearance since I am getting a lot of 60 fps performance - variance 7 % or less. I still see some stutters (partial second pauses) - only in windowed mode - and still have super glassy in FullScreeen. The only change was FFTF=0 and instead of the 30fps next-frame constant stutter with moderate yaw rates, I am getting much closer to pure glass. Maybe there really isn't anything to counteract the occasional stuttering in Windowed mode - or is there? The Fullscreen performance is so satisfying! This testing is with my Q6600/8800GT - I will try to carry over my learning to the 980X/470 system next week.Thanks as always for the amazing knowledge that you have!Dave

Hi i got my rig tested with the tool but a stupid question --now i still have Nhancer in do i take it out?? And something else, when a guy changes some thing in the fsx conf. file when does it take effect , you shut down the sim and restart or how does that work?.And ******* every time i see how devoted you are to this and the patience you have i just shake my head in amazement, i don't understand really what you do, are you changing and rewriting new codes from the original fsx sim from Microsoft?

Hi i got my rig tested with the tool but a stupid question --now i still have Nhancer in do i take it out??

No need to, you can use nVidia Inspector with new drivers (nHancer wont work)

And something else, when a guy changes some thing in the fsx conf. file when does it take effect , you shut down the sim and restart or how does that work?.

Yes, simply make sure FSX is closed. It will simply read the changes when you start it.

And ******* every time i see how devoted you are to this and the patience you have i just shake my head in amazement, i don't understand really what you do, are you changing and rewriting new codes from the original fsx sim from Microsoft?

Nope. I simply look at the code with special tools that help me understand how it works so I can apply tweaks to make it perform or look better.

A question for Hey-sus,First, I also want to express my appreciation for your excellent contributions to this hobby. Now for my problem. I have run your Tweaking and Tuning tool in the past and it ran fine. However, now I get an error message stating: "There was a problem reading your configuration file." The FSX.cfg file is in the default location and I have tried manually directing the utility to the FSX.cfg file with no luck. Thank you for any help.Best regards,Bill

No need to, you can use nVidia Inspector with new drivers (nHancer wont work)Yes, simply make sure FSX is closed. It will simply read the changes when you start it.Nope. I simply look at the code with special tools that help me understand how it works so I can apply tweaks to make it perform or look better.

Hi and Thanks again for all of your great work. You mentioned special tools and it got me to thinking about a problem I have. Like so many others, I use many add ons such as EZDok, FSUIPC, Voxatc and others. All of theses products have both pre-assigned hot keys and keystrokes and the more advanced use special codes and even Macro codes. It is not unusual to find that when you are having trouble with say rudder input to find either a duplicate functions or keys assigned somewhere that are in conflict. Add to this the wonderful gift of FSX trying to add its own pre assigned inputs even after we have deleted them (LOL!) and we have a receipe for disaster. Is there any software that would help in locating and identifying these conflicts. I guess I am thinking of something like the code that seeks out duplicate files or the one that will notify you of missing texture.

I guess I am thinking of something like the code that seeks out duplicate files or the one that will notify you of missing texture.

I don't know of any. However, what I do is un-assign every single key and axis from FSX, assign those I need for my add-ons (using the add-on provided utility) and THEN I assign the FSX keys I need.. ONLY the ones I really need, so I don't experience any conflicts.

I have run your Tweaking and Tuning tool in the past and it ran fine. However, now I get an error message stating: "There was a problem reading your configuration file."

Hi Bill,In the past I would read 'any' fsx.cfg file sent. Now I check the file is strictly ANSI (meaning it contains ONLY ascii codes) so, check that there are no foreign characters on your fsx.cfg. Also, try to open the file and save it again and make sure when you do so it saves as 'ANSI' (you can check the ANSI type when you are about to save the file)

I have a question about the Shader mod.Maybe two months ago i installed this and got an awful look of the water.I have REX installed on FSX ACC.Is this issue meanwhile gone with your Shader mod?Thank youMichael

...Is there any software that would help in locating and identifying these conflicts. I guess I am thinking of something like the code that seeks out duplicate files or the one that will notify you of missing texture.

Have a look here: http://www.fsdevelop..._-_Missing_(FSX)You need to add a line to your fsx.cfg file. Quoted from the FSDeveloper page:

[sCENERY]// To show alerts or areas, set to = 1 // To hide alerts or areas, set to = 0 // Alert for missing texturesShowMissingTextureAlert=1// Alert for missing objectsMissingLibraryAlert=1// Render Trigger areas as visible boxesShowTriggers=1

Hope this helps!

I have a question about the Shader mod.Maybe two months ago i installed this and got an awful look of the water.I have REX installed on FSX ACC.Is this issue meanwhile gone with your Shader mod?Thank youMichael

Michael, there is a new version.. not sure what version you used back then. Check my signature for the link