Best way I have found to - SPEED UP X-PLANE

[captflyby 2]

I build a flight sim, very realistic, using x-plane and simavio2.

(It can also run FSX and Mindstar)

 

I've also read threads in many forums where people have spent a lot of money on computer hardware in hopes of speeding things up, only to be dissappointed with the results.

 

 

i.e. x-plane won't take advantage of multiple video cards, and it does not put hits out of the ball park with multiple cores either. 

 

 

During some recent work with the Mindstar tech support guys, they made me realize that a large portion of the "slowness" involved in a sim was the swapping of scenery from hard drive to memory as you fly.

 

i.e. The best way to speed up your computer hardware for improved sim performance, without upgrading it all, is to upgrade the hard drive.

 

 

Here is what I did to speed up my sim:  This might work or you, it might not - no guarantees.

 

 

I replaced the "C" drive - i.e. the drive that was used by my computer to run x-plane, with a solid state drive.  That took my random seek from 183 to 380. (about $70.00)

 

 

I then bought a 2nd identical solid state drive and raided them together (raid 0).  That took my random seek from 380 to 490 approximately. (It took my son's computer to over 1 gig random seek).

 

 

Now I can set a reasonably complex scenery, using 3rd party mesh hd v3 and other addons, and still obtain a respecable 60-70 fps frame rate. On night flights, it hits 120+ fps.

 

 

The threshhold of reality for frame rate is 60 fps.  i.e. the rate at which the eye/mind begins to perceive the image as it does in real life. 

 

 

So for about $140.00 + tax, and a little sweat equity, my sim now performs about 2 1/2 times faster than it did before.  Smooth motion, responsive, and loads much faster as well.  Should have gone solid state a long time ago!

 

 

Hope this helps.

[CoolP 283]

I can't confirm any relationship of disk performance (tried old HDD, SSD and SSD RAID) to fps. Loading times, so the time to actually start flying or selecting your place, do improve a lot but the fps aren't affected at all.

 

I would like to hear how the XP devs think about it while I would be surprised to read that the fps performance is bound to any kind of disk bottleneck. The worst thing to happen with a really slow disk might be a blurry texture or a 'blank' but you will not, by design, be able to raise fps as your CPU and GPU combo dictates this regime.

 

If there's more to it than anecdotes, I'd be happy to see it. To be fair, this ("disk drives fps") thought came up with other sim engines too, mainly since SSDs are available. Same result.

 

As of 2015, I think most people are able to easily test for themselves (since SSDs are very common), which of course is encouraged. So your thread might serve as a welcome motivation to do so, regardless of the result. :smile:

[captflyby 2]

I can't confirm any relationship of disk performance (tried old HDD, SSD and SSD RAID) to fps. Loading times, so the time to actually start flying or selecting your place, do improve a lot but the fps aren't affected at all.

 

I would like to hear how the XP devs think about it while I would be surprised to read that the fps performance is bound to any kind of disk bottleneck. The worst thing to happen with a really slow disk might be a blurry texture or a 'blank' but you will not, by design, be able to raise fps as your CPU and GPU combo dictates this regime.

 

If there's more to it than anecdotes, I'd be happy to see it. To be fair, this ("disk drives fps") thought came up with other sim engines too, mainly since SSDs are available. Same result.

 

As of 2015, I think most people are able to easily test for themselves (since SSDs are very common), which of course is encouraged. So your thread might serve as a welcome motivation to do so, regardless of the result. :smile:

 

Thanks CoolP,

 

It might very well be as you say, I am just grateful to have the speed I need from the sim at long last - however it got there.

Total b. s.

JSkorna,

 

Sorry you feel that way.

 

Don't BS me, BS the motion picture association of america and their clinical studies.

[scotchegg 4,170]

 

 

Don't BS me, BS the motion picture association of america and their clinical studies.

 

Ah, you must not be familiar with Mr. Skorna. He's not one to be persuaded by any of your fancy 'facts', or 'carefully structured, documented and peer reviewed evidence'. You'd better just apologise for having the temerity to express a position that isn't his somewhere on the internet and be on your way.

[JSkorna 137]

Don't BS me, BS the motion picture association of america and their clinical studies.

 

Can you provide links to the studies that show that computer generated images viewed on a LCD have a threshold of reality for framerates is 60 FPS?

 

Mr. Egg also believes everything he reads on the Internet as true, especially without any type of proof, fancy facts, or carefully structured, documented and peer reviewed evidence.

[Mochi 82]

Does this mean I can now work from home and earn $600 per day? 

 

Finally!

[JSkorna 137]

Yep! If it's written on the Internet, it must be true!!

[captflyby 2]

You guys are hilarious.

________

 

I made my first feature length motion picture at the age of 18, and have been making films/commercials for many years.

 

The study to which I refer, was conducted in association with the New York City College back in the 1970's, long before LCDs and the internet.

 

They hooked up about 100 volunteers to electrodes that would measure their response to the visual cues that they were fed, via a theater movie screen.

 

The purpose was to see at what point the human mind would cease to see "frame" images and with the aid of "persistance of vision" allow the mind

to "see" as thought the film were one continuous action, i.e. real life, not the start/stop/black/start/stop/black of the 24 fps genre of the era.

 

I do applogize for not havign the reference at my fingertips, as this was many years ago, but I have a book on the study somewhere in the clutter.

 

If you happen to have access to an old 8mm movie projector, most of the more advanced ones could be dialed to set the selected fps.  ex. 15 fps, 20 fps, 24 fps and 30 fps - were the most common. 

 

As a youngster, I noticed that if the Charlie Chaplin film I had checked out from the university library was played at 30 fps and not the intended 24 fps, the visual perception of that image

was like watching video tape, not film.

 

Using the same theory, the study advanced the fps until the basic threshhold of reality was fixed at approx. 60 fps.  The reason film was never shot at that speed, (one or two actually were) was that the resulting

film disk, for each reel, would have been in excess of 6 feet in diameter.  Not practical.

 

There is also a reason why the developers of the Oculus Rift have their units set for 90 fps. Not only is the threshold of reality met, but the blurred motion that you get in the theater when

the camera is panned during a sequence (stuttering) is gone.

 

Now to x-plane frame rate.  The lower the frame rate, the more the "stuttering" effect of the visual cue.  This just shoots reality all to ..................andromeda and back.  No sense of really flying

when the outside world looks like a slide show.

 

A friend of mine develops flight sims for different militaries all over the world.  He says, if the visual is good enough to make your hands swet, it's close enough.

 

I cannot speak to those who really know the ins and outs of x-plane, all i can tell  you is that with a simple adding of a solid state drive and raiding two of them, my fps went off the charts.

 

I had previously brought my system to its knees trying to reach the coveted 60 fps threshold.  Now I exceed it.

 

If it happens that I later find that the SSD had nothing to do with my fps boost, then I will post it. But for now...

 

based upon voluminous posts that I have read from others....

 

x-plane is not that favorable to incresed cores

 

x-plane will not take advantage of multiple video cards

 

and many have boosted their hardware to no avail.

 

Doesn't seem like x-plane likes much at all.

 

But it did seem to love my SSD upgrade.

 

Happy simming.

[JSkorna 137]

Grow up. Don't be so damn rude.

OK, total cow pies.

 

And now we resume our scheduled programming.

 

This has been a special report from someone who knows that film studies from the 1970's does not apply to computer screen visuals in 2015.

[vr-pilot 11]

Can you provide links to the studies that show that computer generated images viewed on a LCD have a threshold of reality for framerates is 60 FPS?

 

Here is what NASA (some time ago already) and the FAA studied :

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19880012649.pdf

http://hf.tc.faa.gov/hfds/hfds_pdfs/dot_faa_tc_07_11.pdf

 

According to NASA (1988, page 5) a computed frame update of 30 FPS (smoothness) with a screen refresh rate of 60 Hz (flickerfree) is sufficient for normal applications. For military jet fighter simulation ("to display rapidly changing flight scenery without degradation") they would recommend frame updates up to 225 FPS (for realistically high roll rates of 60°/sec). For pitch and yaw speeds beeing accordingly high, they would even go up to 480 FPS. Although the latter they also concern to be "fortunately somewhat lower"... :blink:

 

Interesting: instead of computing 225 "sharp" frames of 1/225 sec exposure time, NASA here suggests less frames using a longer exposure time (page 6). This kind of natural motion blur (which is somewhere at 1/60 sec exposure time) they call "smearing". I am wondering already for quite a long time why natural motion blur is not incorporated in modern simulation technology. Maybe like NASA already in 1988 states (quote): "It remains to be seen whether it is more difficult to compute a smeared picture or simply increase the update rate." (unquote) B)

 

BTW: the above mentioned FAA article puts itself to an 80 to 100 Hz refresh rate for being flickerfree (leaving out the update frequency subject).

 

Interestingly in the mentioned below article the FAA goes down with the update frequency for HUDs and EFVS displays to only 15 frames per second (lowest):

http://www.faa.gov/documentLibrary/media/Advisory_Circular/AC%2020-167.pdf

 

15 FPS was my lower limit for acceptable FSX performance drops while trying to keep "stable" 20 to 30. Now I am using XP-10 :rolleyes:

[JSkorna 137]

Thanks Claus! So I guess 15-30 FPS is good enough!!

[vr-pilot 11]

So I guess 15-30 FPS is good enough!!

 

Well actually (as of now and today) I would say 60 FPS (with 1/60s exposure time) at 100 Hz refresh rate... B)

 

 

There is also a reason why the developers of the Oculus Rift have their units set for 90 fps.

 

Did they change it to 90 FPS? I thought they were doing 75 FPS, which would nicely fit to the different worlds of 25, 30 and 60 FPS (all being divisible by 25 or 15...)

 

 

the blurred motion that you get in the theater when the camera is panned during a sequence (stuttering) is gone.

 

Because I am (also) from the "motion picture department" dealing every day with shutter effects (especially in CGI), I would like to make clear that "blurred motion" actually reduces stuttering perception in camera pans.

In a real film camera ("old fashion style") the exposure time has to be (mechanically) shorter than the frame rate. The normally 180° exposure sector of the rotating shutter produces a 1/48 sec exposure using a 24 FPS frame rate. This way the blurriness of the motion gets close to the "natural 1/60s" while the frame rate itself is less than half the desired speed. So: 60 FPS at 1/60s would do the trick... ("Avatar 2" was being shot at 48FPS (and in 3D))

 

(EDIT:) one more thought on this: because the closed half of the 180° exposure sector of the rotating shutter is used to transport the film roll to the next frame, the other half of the 1/24 sec "action" gets lost for the human eye resulting in a "residual frame stutter". Using electronic cameras and their shutters, it is possible to have exposure times equal (or greater) the frame rate. This way electronic images of all kinds can have true motion blur without "time holes"... FS running at 60 FPS with "full" 1/60s exposure time would probably look very smooth and convincing.

 

It would really be a good discussion to ask:

are "astronomically" high frame rates of 225 FPS or even 480 FPS (NASA recommendation) for the sake of "ever stutter free" visual perception the right way to go?

Would the use of naturally blurred e.g. 60 FPS (with 1/60s exposure time) at e.g. 120 Hz screen refresh rate not be a better way?

Nowadays computing motion blurs (transforming in-betweens) is quite easy (real time implementation). O.K., they are from the quality level not 100% perfect. But otherwise it would be necessary to compute at least 5 in-between samples resulting in 300 frames to be computed per second which directs to the NASA study:

up to 480 samples per second (equals 8 blurred motion samples per frame in a 60 FPS period) would have to be computed without motion blur, which than would be naturally added by the human eye...

Edited June 26, 2015 by vr-pilot

[Slayer 190]

In simpletons terms the faster the aircraft and or roll rate etc, the higher update and refresh rate are required for a smooth viewing experience.

 

That's why anyone who uses a TrackIR notices the sim bogging down more that someone using a fixed view. I'm no math wiz but it seems the velocity you're moving the viewpoint would dramatically affect this too because it's not just the aircraft in motion your head is as well. 

 

I still fail to see how a SSD will effect your FPS unless your flying at mach 3 . Scenery is cached in ram so it's not continuously paging in from the hard drive. I've been running my sims on SSD ever since they hit the market...

[CoolP 283]

 

 

I am just grateful to have the speed I need from the sim at long last - however it got there.

I would feel the same and I'm happy you took my criticism that openly. :smile: Happy X-Plane flying.

[Smokeyr67 0]

Debate about frame rates aside, my personal experience is that upgrading from a conventional platter style HDD to an SSD will only improve load times and scenery loading pauses.

 

Like putting Cadillac suspension on a Hyundai, It won't go any faster , but it will smooth out the ride:)