FIBER_FRAME_TIME_FRACTION=0.01  In Main Section of Prepar3d.cfg

 

In the past I had 10fps in NY USA. Now, depending upon the view, I get a very acceptable level of performance. All scenery sliders to the right. Limited AI/traffic to 10. MSAA at 2. At 8 =10fps again. 

 

LOW END System/rig/PC

 

nVidia GTX550 1gb memory, Intel 2550K over clocked to 4.3ghz, 8GB of memory/RAM

 

Aircraft P3D/Carenado Bonanza

 

Similar results at the ORBX Redding, CA airport and desert area. I have had some nice FPS in the past but the next day the same area gave me poor results. The tweak is the only change I have made tp P3DV2.4. My clouds are at High Detail 70 miles. Cloud shadows on. but limited to 10,000 meters. First 4 check boxes are set on bothe sides. Water at low. 

 

I have no idea of the underlying processes affected.

 

 

 

 

I think that is the speed at rendering scenery. I might be wrong, which is weird because it should have slowed the speed of loading down.

 

One of the wonders of FSX I guess. It should give you better performance and I guess if you only fly low and slow should not give you an issue with the blurries. I would try something a little faster and report back too to see if you get blurries flying over dense areas like that wonderful shot you posted.

 

I might try this myself for my Airbus, as I dont check the scenery out as much at 30k.

 

Thanks for sharing and looking forward to hearing back!

 

 

I have no idea of the underlying processes affected.

 

Hi Dick:

 

The original reference to "FIBER_FRAME_TIME_FRACTION" ...by ACES' Phil Taylor:

 

http://blogs.msdn.com/b/ptaylor/archive/2007/05/11/tweak-of-the-week.aspx

 

 

 

...and some more  links:

 

https://www.google.com/search?q=FSAX+phil+taylor+FIBER_FRAME_TIME_FRACTION&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a&channel=nts#rls=org.mozilla:en-US:official&channel=nts&q=FSX+phil+taylor+FIBER_FRAME_TIME_FRACTION

 

 

Hope this helps ! :smile:

 

GaryGB

Hi Dick - it does indeed work. In the past with FSX, the downside with using such a low FFTF was almost instant blurries; with the P3D multi-threaded engine, you can have you cake and eat it too

In case anyone is interested in what a "fiber" is:

 

http://en.wikipedia.org/wiki/Fiber_(computer_science)

Hi Dick - it does indeed work. In the past with FSX, the downside with using such a low FFTF was almost instant blurries; with the P3D multi-threaded engine, you can have you cake and eat it too

I have never had FFTF have any impact on FSX or P3d before. But before today,  I have never set this variable below 0.15 in either sim. When I set it to 0.01, in P3d, it does indeed give a few more FPS (maybe 10-15%) and no blurries at all.

I was building up to starting a thread FFTF=0.01 and I am glad it has been started. In my opinion the only .cfg tweak that was worthy of the advent of P3D V2.X was FFTF=0.10. Despite my earlier misinterpretation of FFTF I know get it. So I can say that lower you set it the higher your FPS but the more likely it is that you will see blurries. Given that the the default is 33% it is not surprise that reducing it to zero gives a 50% FPS increase.

However you will see blurries in P3D V2 if you set FFTF=0, However and this is a surprise to me at 0.01 you may not see blurries at all. I am very surprised by this. I thought I was doing well to be able to get the benefits of FFTF=0.10 without blurries but in fact I am getting much better performance with FFTF=0.01 and still without blurries.

 

I will try to quantify this over the next few days but I can tell you that my early impressions are very promising.

I don't think zero works but I will try 0.001 and see what happens:-)

I am surprised I am only hearing about this now

 

so what happened to 0.001?

 

must have melted your PC?

Tested a few FFTF values. I think 0.01 is indeed the best. No blurry's higher frame rate.

0.001 got huge FPS increase but terrible bluries and missing autogen. Graph made by sampling FPS at 30sec intervals on Recorded flight.

 

Strangely enough the Ground textures in the FSDreamTeam where not effected by blurry with a value of 0.001 while the surrounding ORBX ground Textures where a blurry mess. Probably priority in scenery library.

 

0.01 is the winner:-)

 

Here is the flight that was used:

https://drive.google.com/file/d/0B4aX5Q3-OT6kQ0tsTktaTmZtZlk/view?usp=sharing

Nice tests! Just curious, did you test FFTF=0.05 or FFTF=0.005?

Thanks Dave, really interesting result. Two questions:

 

- Which fps limit did you set? According to all I read FFTF does only work with limited fps.

 

- If so, how does this compare (blurry/fps wise) with an unlimited fps Setting (that's what I use at present)?

 

Kind regards, Michael

 

 

Nice tests! Just curious, did you test FFTF=0.05 or FFTF=0.005?

You are correct its a typo on the graph it should read 0.005 i.e. half as much time as 0.01 and virtually no difference that I could see.

 

According to all I read FFTF does only work with limited fps.

Don't think that's accurate. My FPS is set to unlimited in P3D V2.4 Display setting.

Didn't see any blurries until I tried 0.001 and then lots. And as I said no additional benefit to 0.005 but I do think I noticed during the start of the playback on that one an area of ground texture come into focus but just once.

 

I should make a stutters and FPS comparison adjusting the FFTF value with frames locked to 30 and Vsync on.

Thanks Dave. There was, e.g., a thread on the LM forum not long ago with just that general opinion (although no one from the LM staff contributed, as far as I recall). The more I read the more confused I get. Same goes for all those AA settings.

 

Anyway, thanks for the thorough test, very well done.

 

Best regards, Michael

From my experience with "aiming for high FPS" I can only say that a solid balance of parameters will give the most stable results.

IMO the consequence of high FPS settings "here" is creating a bottleneck "there" resulting in stutters "at the end of the flagpole" (as we say in Germany).

Choosing parameters that will result in more than 40 FPS is for sure a bench mark test for high end machines, but here the question might not be if stutters will come, but when they will occur.

So no blurries at 0.01? Interesting. I will also check again.

I will upload the 0.10 and the 0.01 video's to youtube

and post the links here.