Lets talk about Frame-rates.

[Hiflyer 129]

Ok! Looking around, it seems like no matter which flight-Sim forum you visit, you are never far from a question about Frame-rates, stuttering and wildly varying frames per second. Lots of people don't think they have enough, some lucky few are happy, and meanwhile the various forums are full of helpful solutions: set your Buffer-pools to this, set your Affinity to that, lower your sliders a bit, switch to 64bit.......

Its almost as if FSX has become a system tweak utility!

Which has me thinking a bit about this potential new Sim. Will we finally all be freed from the horrible frame-rate monster? And if so, whats the base frame-rate to aim for? I know that on consoles, there is often a set goal of 30 to 60 fps with all settings maxed, and some games increase the apparent frame-rate using tricks like motion blur.....

I know almost everyone must be hoping for the absolute maximum fidelity possible in the new Sim, but I don't think anyone wants to have a tech curve like that of FSX where only now, years later, are we even close to having machines to tame the beast.

I am told that the minimum scenery slider in FSX displays more than the Maximum slider in FS9, and I suspect we will be expecting a similar (or greater!) jump in the new Sim. But at what cost? Is the community willing to wait a few years after release before beginning to see a new Sims full graphic potential, or do the power-users want to "floor it" right from the start, and never have to think the phrase "frame-rate" ever again?

Or is it somewhere in between?

I am just curious about how others might view that question......

[Snave 466]

I have previously explained why fps-at-all-costs is such a red-herring in sim-dom, but also why the perpetual myth of `24 is enough` is actually drivel when you talk about a desktop computer simulation. For a start, the `24 frames-per-second in movies` is a completely false analogy. A piece of film containing a static image passing a light source has NO CORRELATION WHATSOEVER with the calculated, composited, displayed, ADJUSTABLE display of a computer game on a TFT or LCD, and that's before you get into the issue of the display rate of the visual display unit itself. And new digital movie cameras film at 50-60 fps and above... search on ARRI Alexa or view HERE.

It's like comparing the 787 to a Ford Trimotor. Why does the 787 only have two engines, when the Ford had three, so obviously three is better... or could it be that the Ford was 1920's tech and the Boeing is just nigh-on a hundred years later and times have moved on?

FSX, for example, offers better fidelity to background processes when the frame rates are LOCKED. Sorry, not interested in the `smoothness` arguments about running the sim at Unlimited, it achieves those frame rates only by leaving other stuff out...

On the other hand, a minimum requirement also applies - and simple tests bear this out. Lock the fps slider to 20, and you don't get full flight model fidelity. Force feedback effects work (even) less effectively, weather effects become abrupt rather than smooth, incremental and fluid. To name but two.

So the truth is somewhere in between. But if one is starting from fresh, one has the possibility to design the sim from the outset to work optimally at a specific frame rate range. I think you hit the nail on the head: 30-60fps is about right, providing refresh fidelity that allows for system overhead to make the myriad of background calculations for the simulation of flight, not airborne railways on tracks, and coincidentally allows for parity with the 60Hz typical refresh rate of the standard TFT/LCD monitor, whilst not swamping the information channels and allowing for upgrades and advanced addons.

HOWEVER, as 3d is becoming the new `standard` I don't believe there is any way to avoid having a 3d `module` or element in any new sim, and as we all know, current technology for 3d display requires a display frequency that doubles up the monitor Hz rate to 120. I don't see that changing as new monitors are being released to suit this requirement and it's becoming the de facto standard in HDTV. The synergy in designing the display parameters to enable 3d display on a large-screen 3d TV are simply unambiguous, so that is where the display needs to be focussed, if you will.

Whether 3d display can be separated from `per-frame-calculation` for the purposes of flight modelling could be an interesting challenge for the Next Gen Sim developer.

[Rafal Haczek 1612]

I am told that the minimum scenery slider in FSX displays more than the Maximum slider in FS9

I heard that too but my eyes definitely disagree with that. For some time I have been using both platforms (sometimes even launching both on the same day) and have a clear comparison. At least to my visual perception (on the other hand why would I care to pretend having someone else's perception?).

There is also another question that jumps in: what do we mean by MORE? Is it the texture resolution (this is certainly higher in FSX), simply the number of trees or self-repeating houses (not that I care much about it)? Moving cars, boats and birds? Shall we go down to flowers and stones in future? Or maybe just the variety and realism of landscape (that appeals to me more)? I'll leave the answer to each reader.

As for the FPS, my feeling is that it is a neverending story. Pure FS9 on default settings gives you practicaly unlimited FPS. Now move the sliders to the right end, add UTE, GE, FSG, complex aircraft, 100% AI traffic, addon cities and airports, HD clouds, real weather, nav tools, maps, etc - your frames WILL drop. Of course today's hardware will let them still be relatively high. But they will drop, there is no other logical way to go. Graphics must be displayed, animation must be calculated, sounds must be played, etc. Same with FSX. Fill it with all possible addons and tell me your frames stay at the same level as out of the box...

Now, my point is if the next simulators to come offer only the basic sim world (and they will since how many tetrabytes of disc space and how much development labour would be necessary to recreate everything users need?), there will always be market for making addons. And here the story starts again. Fill your new sim with them and FPS will drop and drop and drop.

I'm neither a PC guru nor am I pretending to be one. But let's say (yes, I realize it's simplifying) we need 30 FPS to get a smooth visual impression. So that would mean the limit of stuff we can add to our sims unless we enjoy slideshows. If your FPS start dropping below the 20-30 level, do stop. Then maybe buy a better rig (money money). Then you get a top notch PC and will hear another sim is coming which will make your visual experience deeper. Is there an end to this story? I can't see it.

As some of you know, I am still mainly an FS9 user. The reason being also (but not only!) the performance. Of course I know today FSX can work quite well on good rigs. But much depends on what you like and need in your sim. I am one of the crazy guys who try to recreate the world in their computers (lol, would women ever understand it?). So my FS9 has so many various addons that it performs just right (to the acceptable limits). I am also using FSX now as there are some new FSX-only addons I like. As it is still quite basic in my case, FPS are acceptable (similar to my fully filled FS9). But I can already see how much I would like to add to it to fit my taste and needs. And the story begins all over, doesn't it?

Of course we all know Microsoft Flight SImulator was written not the best way possible and its engine leaves much to wish. Both developers and users can experience that. We can hope the announced Aerosoft Flight Simulator will be way better in this department (I believe it will). But does that mean we won't want (sounds nice) to add sceneries and planes to it? And won't we do it for the price of frames? I don't think it simply is possible.

You know, I am not an FPS maniac. If they stay on a good level, I stop caring if it is 30 or 130. I Really do.

But I need realtive fluidity to enjoy. 'As real as it gets' we say. In real life you don't get low FPS approaching complex objects or don't get vision stutters (another topic). So we do expect to avoid it in our sims.

So, my personal verdict: give me the minimum FPS my eyes need to get fluid display and I'm happy.

[awesome 22]

In my opinion the user demands too much. I mean we want all these add-ons running at max settings. That's unfortunately not always realistic.

Seriously, almost every PC is capable of running FSX at a reasonable frame-rate. Let's take myself as an example: I have a 4 years old single core 2.4 GHz machine and I fly PMDG aircraft. But I still get 30 fps average. As you can imagine my settings are at absolute minimum. But I still enjoy the experience =) Basically what I want to say is that it isn't FSX that creates frame-rate problems, we as the users do.

I am told that the minimum scenery slider in FSX displays more than the Maximum slider in FS9, and I suspect we will be expecting a similar (or greater!) jump in the new Sim. But at what cost? Is the community willing to wait a few years after release before beginning to see a new Sims full graphic potential, or do the power-users want to "floor it" right from the start, and never have to think the phrase "frame-rate" ever again?

Or is it somewhere in between?

I am just curious about how others might view that question......

It think it is going to be something in between because you both want a future-proof sim as well as being able to play it right from the start.

[Snave 466]

Actually, the clue is in the answers in the above, and so is the solution. Sadly something MS abjectly missed in the erroneous `future proofing` concept as they guess at future trends, and failed:

Design the sim to run at near-max settings on TODAYS equipment, and as the sophistication of the addons increases the ability of newly-arrived hardware to run it increases at a commensurate rate provide UPDATED elements to the core sim to allow for increased and new featuresets.

As I postulated in the Aersoft Sim topic, making `Future Sim` in modules, elements that can be individually updated and improved, then slotted into the `mainframe` is possibly the way to go. You don't have to look far for the analogy - we generally all update parts of our computers rather than replace them outright. A new graphic card can double the display fidelity, we can change the hard drives, sound cards and update and increase RAM. All at times to suit our requirements and pockets, in other words over time. Instead of playing the guessing game about future trends, if the sim itself is open to updating, but is designed for the technology prevalent at the time of release then it enjoys not only a longer shelf life, but also a range of setup to suit ALL purchasers, even those who don't update the core modules and gives the opportunity to wait for the positive acceptance of new and derivative technologies over time.

Nex, this `slider adjustment` policy has proved itself time and again to be a mistake.

No matter the intention, simmers try to push the sliders beyond that which is appropriate for their rigs. It's human nature and it cannot be changed. The problem is that they then blame the developer for the fact that there sim isn't smooth, or it blurs, or the sound is intermittent, and excessive tolerance of the ignorance of the whining simmer takes support resources that would be better spent providing genuine support for those that require it, or updating/altering the `faulty` module, or developing its Level 2 replacement while hypothesising about the Level 3 and Level 4 modules and `future thinking`...

Other advantages of this `modular` approach is that it gives aftermarket developers a peg to hang their hats on, with defined limits and integration requirements, yet still permits `out of the box` creative thinking and developments by way of plug-ins and extended modules - another anlogy: All computers and their components plug into standardised internal layouts, using standardised plugs and sockets. The computers and their components plug into standardised wall sockets, use the standardised electrical supply and display on screens that use generic display elements.

But that doesn't deny extrapolated thinking, experimentation and exciting development. It encourages it.

From the commercial standpoint this offers the developer of Future Sim something that MSFS had to buy through the aftermarket, over which they had little control or financial benefit - shelf life. If the sim-developer gets to revisit the sim on a regular basis and sell upgraded, updated components then the profit is theirs. Perhaps this could extend to licensing development to the recognised specialists - why develop advanced weather modules when Active Sky can do it better? Simply license them to do it for you.

This actually also serves the wider Aerosoft business model by providing partnership opportunities that benefit existing developers rather than

marginalising them. It also serves to allow `tuning` of the individual simmers rigs in subtle, external ways to remove the most unreliable, inconsistent and troublemaking element of simming - the end user.

As Mathijs and I know, it takes only seconds to set up even a below-average rig to run FSX in a lucid fashion. Even if it is not the `100% slider` that the simmer impatiently and incompetently demands. Clearly that option for the average simmer is a failure. In fact even for the most experience simmer it's a failure, judged by the number that go straight to the FSX.cfg, Panel.cfg, scenery.cfg et al and make the changes directly when the sliders achieve the exact same result! It is a sad fact that as simulators themselves get more sophisticated, the relative abilities of the customer to take advantage of those increases actually lessens. Future Sim MUST be designed to suit mass-market and niche-market alike, even if that means making huge numbers of placebo adjustments. Modules would provide a controlled method of eliminating simmer stupidity and could even be tied to individual hardware componentry. ATI-optimised graphic module, NVidia-optimised graphic module, SSD-qualified data transfer etc. etc. And allows for the Release Version to be suited for the hardware prevalent at the moment of release, instead of over-reaching and miss-speculating on tomorrows tech. Which means the sim runs better on more computers, more of the time, with happier customers.

Fps and smoothness are key elements to this, and to design anything that will only run smoothly after the purchaser has made hardware upgrades is totally self-defeating.

So here's my vision of the future:

Saturday, the Aerosoft Flight Sim roadshow comes to town to the reseller near you. You and your son/father roll up to see what's on offer, as do all the other customers at the store to buy new hard drives, extra RAM or that new whizzbang I-Pod. Instore a Huge Screen 3d demo unit has been set up - using proprietary hardware that is within reach of even the average customer. You sit the potential customer down and they hand fly a circuit of an airfield in a Cub or land a Jumbo at a busy international hub, in 3d, with surround-sound. Aided by qualified personnel and the choice of the `basic` or `advanced` aerodynamics module, the seven or seventy-seven year old can get the satisfaction of a great landing. Hey, they're a PILOT! And then you show them - and all the others tha have gathered round ni their 3d glasses - that they can have this sim on their 3d TV at home, today, using the hardware they already own. WYSiWYG.

Then, to impress the natives still further, you switch to the High-Fidelity Level 2 modules which contain the latest developments and now that circuit in the Cub has to avoid low-flying birds, while the weather causes the low-level turbulence of thermals from the fields and housing estates around the field, all while watching the engine vibrate in front of you. Or the Jumbo now has a Co-pilot calling out the checklist and actually setting the flaps and power for you while advanced ATC talks to the fifty other aircraft stacking up behind you and traffic crosses the runway on the airport in front of you. Wow! now it's not just good, it's as real as it gets (to coin a phrase...)

And yes, I deliberately choose things which are here, now for FSX. I have ideas for other modules, but that's my secret...

Perhaps a sample addon-module or two to show how they add to the basic sim over time, or can set the sim AUTOMATICALLY to take advantage of the latest features of Ati or Nvidia cards even as they improve their home entertainment or PC system. At the end of the demo, when you ask for the 3d glasses back, you simply ask which version sir or madam would like for themselves or little Johnny...

Now THAT'S an assumptive sale, supported and created by future-sale benefits. You've future-proofed by ensuring that the customer develops with you but without relying on their knowledge and expertise to do so and without risking the sale to do so. I previously used a similar model in a different industry (`turn up, try it, buy it, add to it`) and made the brand iconic.

Simple.

[bionicCrab 7]

Developer of Future Sim won't get it right because It'll be microsoft again.

[Gaura Mohana 23]

Developer of Future Sim won't get it right because It'll be microsoft again.

didnt you hear?

Microsoft closed ACES, so we wont be seeing any more flight sims from them...

now Aerosoft, and a few other companies are the only ones making (or atleast planning) a new flightsim

[bionicCrab 7]

What does Aces have to do with Microsoft Flight simulator? Nothing anymore....in fact microsoft stripped Aces of the flight sim engine months before they fired everybody. I don't know but normally if a title makes you excessive profits you don't get rid of it... and microsoft NEVER said anything about flight simulator, they only ever cancelled train simulator 2 officially. And besides FS Live is now by Microsoft Game Studios, bottom line they were getting rid of dead weight.. evidenced by such wonderful startups as "cascade game foundary"..

[Hiflyer 129]

Actually, the clue is in the answers in the above, and so is the solution. Sadly something MS abjectly missed in the erroneous `future proofing` concept as they guess at future trends, and failed:

Design the sim to run at near-max settings on TODAYS equipment, and as the sophistication of the addons increases the ability of newly-arrived hardware to run it increases at a commensurate rate provide UPDATED elements to the core sim to allow for increased and new featuresets.

Pretty much parallels my own gut hunch, but I was not sure how many might also feel that way. For the tech-heads amongst us, there is a certain sexiness in hearing that a game is so advanced that nobody can run it. Its kind of a direct challenge! (Think of Crysis)

Of course, many other people are simply excluded. Perhaps permanently.

As I postulated in the Aersoft Sim topic, making `Future Sim` in modules, elements that can be individually updated and improved, then slotted into the `mainframe` is possibly the way to go. You don't have to look far for the analogy - we generally all update parts of our computers rather than replace them outright. A new graphic card can double the display fidelity, we can change the hard drives, sound cards and update and increase RAM. All at times to suit our requirements and pockets, in other words over time. Instead of playing the guessing game about future trends, if the sim itself is open to updating, but is designed for the technology prevalent at the time of release then it enjoys not only a longer shelf life, but also a range of setup to suit ALL purchasers, even those who don't update the core modules and gives the opportunity to wait for the positive acceptance of new and derivative technologies over time.

All very juicy. But also very, very ambitious.... Time will tell if Aerosoft is setting more cautious, modest goals, or really is full-out betting the farm!

Nex, this `slider adjustment` policy has proved itself time and again to be a mistake.

No matter the intention, simmers try to push the sliders beyond that which is appropriate for their rigs. It's human nature and it cannot be changed. The problem is that they then blame the developer for the fact that there sim isn't smooth, or it blurs, or the sound is intermittent, and excessive tolerance of the ignorance of the whining simmer takes support resources that would be better spent providing genuine support for those that require it, or updating/altering the `faulty` module, or developing its Level 2 replacement while hypothesising about the Level 3 and Level 4 modules and `future thinking`...

I am not so sure about eliminating sliders...... I have actually used software that attempted to automatically gauge my systems capability's and set the graphics options accordingly. It turns out that the best thing about that was that I could override it!

Part of the problem is that these things tend to work better for middle-of-the-road machines and non-technical users. Those capable of tweaking their machines innards and who tend to be always on the bleeding edge of performance find that the more over the top systems are only rarely taken into account by those settings, and so you wind up with something far below the machines actual capability. If I was stuck with that, with no options, I would be quite....... perturbed.

[Snave 466]

What does Aces have to do with Microsoft Flight simulator? Nothing anymore....in fact microsoft stripped Aces of the flight sim engine months before they fired everybody. I don't know but normally if a title makes you excessive profits you don't get rid of it... and microsoft NEVER said anything about flight simulator, they only ever cancelled train simulator 2 officially. And besides FS Live is now by Microsoft Game Studios, bottom line they were getting rid of dead weight.. evidenced by such wonderful startups as "cascade game foundary"..

Er, I dunno where you got that info from but it's totally duff, mate. MS didn't just lock the MSFS franchise in the vault alongside the Train Simulator 2 project, they also off-licensed the ESP project to Lockheed Martin, further rendering your analysis off-kilter. Where you are correct is that Microsoft have not ruled out future flight sims...

Official statement HERE, however that preceded the licesning of ESP by some ten months. At the moment there is no ongoing flight simulator project at MS. Not for the PC, not for any game engine. Committed to support for the current product is a mercilessly long way from commitment to future development of the genre.

So you can see they DID say plenty about Flight Simulator, but what they SAID and what they DID are, as often, two completely different things.

Remember though, that MS also has NO history of instigating flight simulator development. Just as they happily bought out Bruce Artwick 26 years ago, they could quite happily buy out any other developer that uses a Windows-based, DX11 developed platform for a Future Sim. Which rules out Austin and X-Plane and all of the existing or known general flight simulation products.

My guess - and it is purely that - is that te license deal with Lockheed Martin is designed to keep the platform development alive, with MS able to buy back in for the non-licensed commercial platform market. Less of a guess is that when LM find out what a mush of code ESP is, they will quietly put it on a shelf somewhere and let it gather dust. ESP cannot deliver a DX10-qualified product so there is no hope whatsoever of a DX11-optimised new build anytime in the near future.

But I do agree that we should never say never.

[bionicCrab 7]

Really then how come Microsoft is actually hiring for FS Live? And licensing a code base is not quite giving away the entire franchise..

[Hiflyer 129]

Actually, the impetus for this thread came partially from an article I found that was discussing the gap between FS9 and FSX. The Author was saying that one of the main reasons so many people were still sticking with FS9 so many years after FSX was released was that while FSX had great potential, it was not potential that could be realized by most peoples machines, especially at the time.

Rather than eagerly making the jump to the next generation, many hung back, slowing the uptake of the newer Sim. The article then went on to look at forum traffic on the various Sim websites in the years after the release of FSX, and concluded that there had been a fairly sharp decline.

Which made me wonder (assuming that all of this had some factual underpinning) how the community might react if a new Sim ever arrived that was once again technologically out of range of the majority........?

It seemed an interesting thought experiment.

[Snave 466]

Really then how come Microsoft is actually hiring for FS Live? And licensing a code base is not quite giving away the entire franchise..

Really? And how do you know that FS Live is a replacement for the MSFS franchise and not a flying game fps shoot-em-up..? I can buy a Mini, paint `Ferrari` on the side and colour it red, but that don't make it a Ferrari...

[Snave 466]

Actually, the impetus for this thread came partially from an article I found that was discussing the gap between FS9 and FSX. The Author was saying that one of the main reasons so many people were still sticking with FS9 so many years after FSX was released was that while FSX had great potential, it was not potential that could be realized by most peoples machines, especially at the time.

Rather than eagerly making the jump to the next generation, many hung back, slowing the uptake of the newer Sim. The article then went on to look at forum traffic on the various Sim websites in the years after the release of FSX, and concluded that there had been a fairly sharp decline.

Which made me wonder (assuming that all of this had some factual underpinning) how the community might react if a new Sim ever arrived that was once again technologically out of range of the majority........?

It seemed an interesting thought experiment.

I would be loathe to blame fps for the relative failure of FSX, although I can't see it wasn't a factor. Bear in mind the `over-reach` by MS wasn't just FSX, it was launched as the de facto Flagship of the all-new Vista OS, replete with DX10. It didn't actually turn out that way in reality, and for all the technical and hardware development in the intervening period the path trodden by FSX has been shown to be a blind alley. How much of that is due to FSX, and how much due to the relative failure of Vista and the appearance of Win 7/DX11 is subject to conjecture but taking the fps argument in isolation as the reason for the FSX debacle is probably missing other serious causation factors.

One does wonder, given the glimpse given by DX10 `Preview` what might have been achieved in a fully-optimised Service Pack 3, but we were never know. Even prior to be given the boot, ACES had turned their back on FSX and were already laying the groundwork for FSXI. And this was ALSO a repetition of the old days - as anyone who recalls the FS98/FS2000 transition will remember. FS2002 was the sim that FS2000 ought to have been. Perhaps FSXI would have been the sim we yearned for with FSX, but the point is moot.

According to ACES FSX was designed to work optimally at lower than FS9 frame rates. But the they would say that, wouldn't they? What they OUGHT to have done is simply remove the fps counter in its entirety, along with any internal method of displaying same, but they didn't, so simmers keep on using it as a yardstick even though we know that the act of displaying the on-screen fps skews the result, and no-one seems sure about what that fps display actually shows - spot rate? spot rate over consecutive frames? mean, mode or median average? When one adds in the average frame rate measure manually one gets a far better assessment of the performance of the rig, but of course at a lower persistent average than the glorified spot rate. Which would only have made things worse, and would have caught out ACES in the lie of `lower frame rates are what we strove for...`.

The flight sim genre does not require humungous fps to be successful, but it does require a certain level of fidelity to be credible, and that does require a certain level of frame update. So in part, FSX is a victim of fps, but I don't think sole cause of its demise.

[bionicCrab 7]

Simon; do you have any formal computer science training. Some of the things you say are quite wrong... in fact they're just made of pretenses you come up with to justify your own points.

First of all how is FPS counter inaccurate? Any errors it has are systematic, in fact the fps is a moving average. Second of all Medians and Modes are not measures of the average they are measures of central tendancy! So unless you have a distribution for FPS that we all don't know about, those measures are completely useless.

So i'm guessing.. you've never taken a single statistics class now have you? Otherwise you'd know what those words actually meant.

[Snave 466]

Simon; do you have any formal computer science training. Some of the things you say are quite wrong... in fact they're just made of pretenses you come up with to justify your own points.

First of all how is FPS counter inaccurate? Any errors it has are systematic, in fact the fps is a moving average. Second of all Medians and Modes are not measures of the average they are measures of central tendancy! So unless you have a distribution for FPS that we all don't know about, those measures are completely useless.

So i'm guessing.. you've never taken a single statistics class now have you? Otherwise you'd know what those words actually meant.

Nice attempt at deflection. If you can't attack the message, attack the messenger. So you DON'T have any information on FS Live and were just throwing out your own personal supposition based on what you think you've read? Super, you're entitled to your opinion. Even when you can't justify it. If you have evidence that the much-mentioned, never-seen FS Live is anything other than a first person game rather than a simualtion product, please feel free to present it. Right now.

As to the frame rate counter, moving average of what? If frames are 24 per second, what frequency does the fps counter update at? Does it report the previous second? Two seconds? One fifth of a second, or does it extrapolate how long the last single frame took to display, and multiply the time function to offer a frame rate? If so, what's the duration of the measure period? And what is the duration of the display period? after all,it wouldn't do for it to be reporting the frame rate from ten seconds ago, or extrapolating what the next second or two will bring in terms of display rate, would it? And that's before we consider the rate of monitor refresh, inducing lead/lag in the system.

And on that basis, central tendency would appear to be not only far from useless, but also extremely relevant. Wouldn't you agree, mathematically speaking? Otherwise it might actually be a measure of standard deviation. Oh wait! That's what is measured by the Average frame rate counter that can be manually added to the .cfg file...

..or is it?

[Robert S 75]

So you DON'T have any information on FS Live and were just throwing out your own personal supposition based on what you think you've read? Super, you're entitled to your opinion. Even when you can't justify it. If you have evidence that the much-mentioned, never-seen FS Live is anything other than a first person game rather than a simualtion product, please feel free to present it. Right now.

Simple and easy to see Simon, just follow this link (Note the title of the job - it's for Flight Sim Live)

https://careers.microsoft.com/JobDetails.aspx?ss=&pg=0&so=&rw=15&jid=14542&jlang=EN

Also note in the qualifications: "Flight Simulator or real world flight experience is a bonus. Passion for learning about flight is a must!" They don't mention Flight Combat Sim experience though...but there is the emphasis on "passion for learning about flight".

Nice attempt at deflection. If you can't attack the message, attack the messenger.

Simon - Live by the sword,....

[bionicCrab 7]

Snave: once again you have proven us all wrong...I had not thought of all those possibilities simply because NONE of them relate to the problem at hand. I find you amusing none the less...

Did it ever occur to you to count the number of frames and then divide by the delta time from beginning of the simulation to the end. Or just use the last 15 or so seconds? Or that animation/physics works not by fixed frame rates but delta times as well?

[Gaura Mohana 23]

Excuse my ingorance, but what is Delta-Time?

[OPabst 2091]

Snave: once again you have proven us all wrong...I had not thought of all those possibilities simply because NONE of them relate to the problem at hand. I find you amusing none the less...

Did it ever occur to you to count the number of frames and then divide by the delta time from beginning of the simulation to the end. Or just use the last 15 or so seconds? Or that animation/physics works not by fixed frame rates but delta times as well?

Sorry, but the "Frame counts" you see at the FS (as long as you don't change the Frame Display to the advanced view), is nice to have and says nothing, only a counter how often the engine runs last second. But, that's a less important, because the engine can run 50 times in the first half of the last second and only once in the rest, then you have nice frames counts, but a bad view. The advance view will give you much more info's, like min, average and max frames over the last time (I have no info how long the base of the average is, maybe one minute).

And, there you also see a V: x.y % value at the end of the display line, this is a much more informative indicator about the quality of your "Frames". When this % Value is small, the engine is running smooth (time for a engine cycle is in the same range and has a small variation).

And so, 20 frames with a very constant cycle time is much more smooth, then a 90 frames where the it "jumps" around.

[Sharrow 107]

To enable the advanced frame rate view as Oliver mentions above you need to add the following to your fsx.cfg file:

[TextInfo.2]

AverageFrameRate=1,3

[TextInfo.3]

AverageFrameRate=2,2

//LockedFrameRate=2,2 (This is originaly there - just comment it out with the //)

Please correct me if I missed something as I am writing this from memory.

[Snave 466]

Yes, that's about it. The textinfo.x simply tells FSX which of the three `pages` to display the average display on. So you can place it on one, two or all three of the pages. Generally, one is enough as the DISPLAY ITSELF HAS AN FPS IMPACT.

This then creates a far more useful tool to measure frame rate, even though the `spot` figure, as Oliver states is actually an inherently inaccurate figure. For the simmer, the goal is consistency and smooothneSss, what these additions provide is a better result indicator for any tweaks to system or sim.

In simple terms, the Ave. Frame Rate indicator provides Max frame, Min Frame, Ave frame, and - oh deary, deary me, the percentage in central deviation of fps over the last sixty seconds, or however long the window has been active, whichever is the less. It's the percentage indicator that is the useful tool, although used in conjunction with the `last 60 second average` indicator. What you are trying to achieve is to make that percentage variation as small as possible. So you set the sim, the computer, your tweaks, snake oil or sacrificial goat to get the average to vary as little as possible by setting the fps LIMITER to a setting that is within a fps or two of the average frame rate counter, usually a couple under, and that should bring the variation percentage down to a few small points. And that is GOOD for FSX...

This is the ONLY time I would recommend the use of the UNLIMITED setting in the fps slider as it enables the establishment of the baseline. If you then use it for simming, and open the ave. frame rate counter you will see the percentage variation varying excessively and the average fps going up and down like a whores drawers. Which conforms to the theory that FSX is NOT then showing or calculating all the possible calculations in the time frame and is in fact dumping elements to try to maintain the excessively high frame rate.

But that's another story.

[Snave 466]

Sorry, but the "Frame counts" you see at the FS (as long as you don't change the Frame Display to the advanced view), is nice to have and says nothing, only a counter how often the engine runs last second. But, that's a less important, because the engine can run 50 times in the first half of the last second and only once in the rest, then you have nice frames counts, but a bad view. The advance view will give you much more info's, like min, average and max frames over the last time (I have no info how long the base of the average is, maybe one minute).

And, there you also see a V: x.y % value at the end of the display line, this is a much more informative indicator about the quality of your "Frames". When this % Value is small, the engine is running smooth (time for a engine cycle is in the same range and has a small variation).

And so, 20 frames with a very constant cycle time is much more smooth, then a 90 frames where the it "jumps" around.

Thank you Oliver, my point. Exactly.

[bionicCrab 7]

I've said my peice.. oliver.. please understand the question before you even attempt to give an answer. We're not talking about anything you just mentioned. Especially not graphics issues where you get 50 frames in half a second and 0 the other. that's a resource allocation error, not a problem in timing frames.

Every engine counts the number of frames and the time past since the start of the simulator! I can't see how you can even argue against that point. And further more if you do fixed rate animation you can never recover from that error you describe... the method I described does.

[Paul Golding 159]

Excuse my ingorance, but what is Delta-Time?

Complete explanation follows...

--------------------------------------------------------------------------------

In this discussion, as it relates to kinematics and the study of physics in general, the Greek letter delta means 'change in'.

The Greek letter delta, when used this way, looks like this:

It means 'change in'. So, if the variable x stands for position, then this notation:

is read 'change in position'. It could also be read 'change in x'.

If the variable v stands for velocity, then the following notation stands for 'change in velocity', or 'change in v':

And 'change in time' would look like this:

--------------------------------------------------------------------------------

Now, any time there is a change in a quantity that change is calculated by taking the later value for that quantity and subtracting from it the earlier value for that quantity. So, for example, if an object were first at a position given by the coordinate 3 meters and later at a position given by the coordinate 8 meters, then its change in position would be calculated this way:

The first position at 3 meters would be called x1, and the second position at 8 meters would be given by x2. So, the complete notation demonstrating the calculation for the change in position would look like this:

Therefore, we would say that the change in position was equal to 5 meters.

--------------------------------------------------------------------------------

Note that changes in a quantity can be negative. If an object begins its movement at a coordinate of 9 meters and ends its movement at a coordinate of 3 meters, then the calculation for its change in position would look like this:

--------------------------------------------------------------------------------

Our latest examples here have been with changes in position, but delta calculation proceed the same way with other quantities. Here's another example. If an object was moving with a velocity of 8.2 m/s originally and then changed its velocity to 3.6 m/s later, the calculation for the change in velocity would proceed this way:

And, if an experiment started when time was 3.5 seconds and ended when time was 12.9 seconds, then this would be the calculation for the change in time, or the time period, for the experiment:

One note about the calculation for delta time: Unless you are working on some special time reversal problem, (which would be unusual), if you end up with a negative change in time, very most likely you have made a mistake.

--------------------------------------------------------------------------------

So, in general, all delta quantity calculations proceed the same way. Take the later quantity and subtract from it the former quantity. Here, we show the calculation with an abstract quantity, 'it'.

--------------------------------------------------------------------------------

Hopefully, even if you are unsure regarding the meaning of the following symbols, the delta calculations make sense:

--------------------------------------------------------------------------------

In all of the above examples the subscript 1 and the subscript 2 were used to designate former and later respectively. The former value, (sub 1), is often called the original or initial value. And the later value, (sub 2), is often called the final value.

If we let the subscript 'o' represent the original value and let the subscript 'f' represent the final value, then the following notation can also be used for the change in position:

The above would be read: 'The change in position is equal to the final position minus the original position.' Or, it could be read as: 'Delta x equals x final minus x original.'

Here is similar notation for a change in time:

--------------------------------------------------------------------------------

One can also use the subscript 'i', which means initial. Initial in this context is synonymous with original. The above notation for delta time could be also written as:

This would be read: 'The change in time is equal to the final time minus the initial time.' Or: 'Delta t equals t final minus t initial.

--------------------------------------------------------------------------------

Basically, '1' and '2' form one group of subscripts. Another group is 'f', 'o', and 'i'. One should not mix notation from one group with another, as in:

--------------------------------------------------------------------------------

Using position, or x, as an example, notice that by using algebra we can rearrange any of the above delta equations so that it solves for the first or the second value of the quantity. Specifically, if:

Then we can solve for x2:

The above can be read: 'The second position equals the first position plus the change in position.' But, in general, regarding a delta equation for any quantity, think of it this way: 'The amount of a quantity that you end up with is the amount of that quantity that you started with plus the amount by which that quantity changed.' Certainly one can follow that the amount you have in your bank account after a deposit is equal to the amount you had before the deposit plus the amount of the deposit, and the above works the same with position coordinates.

And we can solve for x1:

This could be read: 'The first position is equal to the second position minus the change in position.' But, again, see this equation as it might apply to any quantity and also notice: 'The amount of a quantity that you started out with is equal to the amount of the quantity that you ended up with minus the amount of change which that quantity experienced.'

--------------------------------------------------------------------------------

So, in summary:

The Greek letter delta, , means 'change in'.

This 'change in' refers to a change in some quantity.

The change is calculated by subtracting the earlier value for the quantity from the later value of the quantity.

This subtraction may yield a positive or a negative result. It may yield a zero change result.

You can use the subscripts '1' and '2' to refer to the first and second values respectively.

You can use the subscripts 'f', 'o', and 'i' also. The subscript 'f' means final, and the subscripts 'o' and 'I' mean original and initial respectively.

Also, closing note:

The Greek letter delta can appear in other contexts and have other meanings. It can appear upper case or lower case, written as print or script. For example, lower case script delta is often used as a symbol for an angle, and in advanced calculus lower case script delta is used in the symbol for a partial derivative. But when delta is used as explained above, (in upper case print), it always means 'change in'.

And no, I'm not that smart; I Googled it like you really need to do.

http://id.mind.net/~zona/mstm/physics/mechanics/kinematics/whatAreDeltas/whatAreDeltas.html