FsDreamTeam to support Lockheed Martin's Prepar3D

[virtuali]

We are pleased to announce we are now supporting Prepar3D with our products, starting with the recently released Los Angeles International Airport (KLAX).

We would like to hear your feedback about how the installation works for you so, we put a test version of the KLAX installer here:

http://www.mediafire.com/?kxcdrr5842vai3w

The installer works in FSX and Prepar3D too. If Prepar3D is found in your system, you will be asked if you want to install KLAX there, otherwise it will install in FSX (if you have it installed).

As with all our products, the installer is a Trial version of the scenery, and you can purchase it either from inside Prepar3D, or on our Web page.

The good news is: we are not requiring an additional license for Prepar3D, if you already own a valid KLAX Serial Number from your past FSX purchase, you can use it to activate the Prepar3D version.

If you run both FSX and Prepar3D on the same machine, and your KLAX was already activated under FSX, there will be no need to do anything: the Prepar3D version of the scenery will work as a fully registered version.

If you haven't purchased KLAX yet, the price it's the SAME as the FSX version and, by purchasing it, you'll get the right to use it under both FSX and Prepar3D.

Please, let's give us some feedback about this version, so we can go ahead and update all our other installers too.

NOTE: after you install the scenery, you'll likely have Prepar3D crashing on exit, because there are two additional menus added, and there seems to be a problem with the current Prepar3D version that any addons that add custom menus will cause a crash on exit, LM people are aware of this and will surely fix it. Also, the scenery works only with Prepar3D 1.2.3253, which is the current version.

[pride545]

How can u tell the difference between the Prepared and FSDT version?

[Bruce Hamilton]

The FSDT version?  You're joking, right?

[virtuali]

How can u tell the difference between the Prepared and FSDT version?

As I've said:

The installer works in FSX and Prepar3D too. If Prepar3D is found in your system, you will be asked if you want to install KLAX there, otherwise it will install in FSX (if you have it installed).

The scenery is exactly the same, the only difference is the Addon Manager and Couatl which have been updated to support Prepar3D too, and the installer, which has been updated to recognize the Prepar3D folders locations.

When'll we have enough feedback, the FSX/Prepar3D-aware installer will be the only installer we'll have, for all sceneries.

[pride545]

The FSDT version?  You're joking, right?

First, you need to understand the question. Then make your comments!

[cmpbllsjc]

The FSDT version?  You're joking, right?

First, you need to understand the question. Then make your comments!

Sarg, I think the reason Bruce was joking with you is because of what your post said, ie. How can you tell difference between P3D version and FSDT version? Where as I imagine you meant to say the difference between the P3D version and the FSX version.




Umberto, since you have tested the scenery in P3D that you wrote the new installer for, on your computer have you noticed a marked improvement in performance using it in P3D on your test computer vs. FSX on the same rig?

[virtuali]

Umberto, since you have tested the scenery in P3D that you wrote the new installer for, on your computer have you noticed a marked improvement in performance using it in P3D on your test computer vs. FSX on the same rig?

I can't see much difference, probably because our sceneries are already optimized a lot. I've tried to match Prepar3d.cfg and fsx.cfg settings one by one manually, but really can't say one is faster. I haven't tested much in other places, though, maybe they worked on default sceneries so, people comparing on a default scenery might notice improvements. In fact, most of the .BGL, especially related to the Cities in the default Scenery folders, are way smaller in Prepar3d, so I'd say they worked mostly on the infamous "frame rate pits", which are default big cities in FSX.

All those reports of great increase in performances, are probably due either because people installed Prepar3D from scratch, so they started with a clean system and/or a new Prepar3d.cfg OR because it's possible that Prepar3d added fixes that optimized things that were slow in FSX, like FS8 ground polygons, which make sense, since you usually start by optimizing the slowest things first.

[Highiron]

I've installed KLAX into P3D and it's been working wonderfully once we got the bugs worked out.

Prepar3D does have some modified code that uses system resources better, such a memory, so there is improvement in the overall performance of the platform, although the frame rates remain about the same. Not all about a platform can be judged by it's FPS alone though, so one must be cautious about using it as the end all be all of performance. It seems other things besides the graphics are being handled much more smoothly in the background. The fact that after several hours of use I have not had a single CTD or freeze, or that it's "stopped responding and needs to close" is somewhat of a testament to it's stability. Something I could not achieve with the FSX Gold Edition right out of the box. Five minutes into use and the program CTD'd without a single add-on present. Only after tweeking the FSX.cfg file would the platform run properly and still every now and then it gets temperamental and locks up or stops responding.

After fiddling around modifying some aircraft.cfg and .air files, I'm also inclined to believe there has been moderate improvement on the overall way it handles aerodynamic theory. When running modified aircraft files in both P3D and FSX, I am not getting the same results. The FSX files need "cheats" in order to behave properly, numerical values that are not true to the aircraft. In P3D I am able to use the real values and get correct results. Using the real values in FSX has resulted in the aircraft being barely able to get airborne, so there seems to be more going on under the hood or bonnet of P3D than a simple, "is there a FPS increase" over FSX. In my opinion the Prepar3D platform is handling processes much more efficiently than FSX. So to say that Prepar3D is FSX re-branded would be incorrect in my belief. Lockheed Martin has had this in their possession for several years now, and clearly they have not sat on their collective hands with it. I believe folks need to step back from the holy "FPS" grail when it comes to determining how good a platform any sim is. Humans can't detect anything above 24 FPS anyway, so why the focus on wether it can achieve greater numbers than any other available or soon to be available platform. The only value it serves is how well it's handling the graphic portion of the engine, but what about the rest of the very important calculations that make up what a flight simulator does or is expected to do? Most of the improvement I am finding in P3D are in the realms the casual user can not see, so hence concludes there is no difference, no improvement. This could not be further from the truth in my experience so far with P3D. The fact is, it is doing things better, you just can't see it...

I eagerly await updates to the installers of other FSDT products into P3D and am more than happy to run the trials on them and report issues if any.

[virtuali]

I believe folks need to step back from the holy "FPS" grail when it comes to determining how good a platform any sim is. Humans can't detect anything above 24 FPS anyway

I agree to everything you said, except this.

That humans can't detect more than 24 fps, is wrong, the 24 fps is the standard frame rate used in movies, but doesn't have anything to do with the human ability to detect motion, it was just the most practical compromise between the cost of the film, the ease of manufacturing camera/projectors mechanics and a "good enough" smoothness.

But, most of all, it's wrong to compare movies/tv to computer generated graphics and assuming that, if movies/tv can do with 24/25/30 fps, the human eye can't detect more. It doesn't work like that.

movies/tv have MOTION BLUR, because cameras have a shutter that is kept open for a finite amount of time capturing the continuous"analog" reality and, the more the shutter stays open, the more motion blurred the filmed frame will be. Typical shutter times used in movies are 1/48th second, for reasons related to the projectors mechanics.

Motion blur is used by our BRAIN (which has a "frame rate" detection capability WAY higher than the eye) to reconstruct the real world fluidity with as few as 24 fps. Loot at a video shot with a very fast shutter time, like 1/1000th of a second, and it will appear to be far less smooth, even at the *same* 24 fps. Some directors used it for special effects only, but it's never a standard, and it's easy to spot when it's used.

Graphic generated on a PC in real-time, doesn't have any motion blur, as if it was shoot with an "infinite fast" shutter time so, it needs MORE fps than tv/movie to appear smooth, I'd say 60 it's the minimum. In fact, professional flight sims needs to offer 60 fps to get certification for training, this wouldn't be necessary if the "human eye couldn't detect more than 24 fps"...

So no, higher than 24 fps in a PC sim ARE desirable, because realtime PC graphics it's a different thing than tv/movies.

On top of that, when judging a sim, there are several other factors that comes into play to give the illusion of reality and smoothness, which are:

- Frame rate consistency - saying 30 fps doesn't mean anything more than "30 frame per second", it doesn't guarantee that EVERY frame is always happening *exactly* every 1/30th of a second. Having 30 frames generated in the first half of a second, and then a pause for the other half is STILL 30 fps, but it's horribly unusable!

- Response to user controls input - even if the visual system is updating at 60 fps, it doesn't help if the user inputs are sampled at, let's say, a 10 hz, because it would result in a 1/10th of a second lag between your input and the result on the simulated image. This also affect your feel of reality, a lot.

- Flight model updates - even if the visual system AND the user inputs were processed at an high rates, it wouldn't help if the flight model calculation weren't following. If you are reading 60 fps on screen, but the flight model is updating at 30 hz, you ARE seeing 30 fps anyway, because video frames generated in the middle of a flight model update (the airplane position and thus your point of view hasn't changed), are just copies of the previous one, so they don't contribute to the illusion of smooth movement.

So yes, I'm all for not being obsessed with the visual fps indicator, because it doesn't really tell the whole story, and it's entirely possible that Prepar3d might "feel" smoother even at the same fps, because changes in the underlying code allowing this are definitely plausible (FSX itself felt smoother than FS9, at the same fps), but saying having more than 24 fps is not useful, because the human eye can't detect more, it's one of the most widespread myths that needs to go away...

More info about the subject here:

http://www.tweakguides.com/Graphics_5.html

[AaronMyers]

Yeah, the statement that "the human eye can only detect <insert arbitrary number here>" is probably the most misused statement in forums today beating out the other highly popular "it looks cartoonish". The ability to discern frame rates higher than 24 is easily tested by simply locking your frames at 24, test, then raising them to 50 and test again. Unless you have a vision correction issue, or your system is unable to reach a level higher than 24, you will see the difference.

[Highiron]

100% of your reply is correct Umberto.

I was being very general about the 24 FPS human limit, and yes I certainly had the film entertainment industry in mind with the statement, a poorly formed example on my part. My error. A more appropriate statement would have been, at what point does the FPS become a problem on a simulation platform? An example would be the 24 FPS human limit with film. The speed of the action determines the minimum frame rate that stuttering is percieved. You can not tell the difference at 24 FPS vs 50 FPS cruising at altitude in straight and level flight. Close to the gound, yes, and if your tuning than even more so, and on top of that by the rate of turn. If you wanted to do aerobatics perfectly smooth, you would probably be in the 100+ range. The question is, is it a realistic expectation with a home based system and the massive eye candy add-ons we throw at it and the normal use of a flight simulator?  Most frame limiters default to 30, this is no random number, it's a compromise.

The human eye/brain interface is extremely complex and is capable of seeing a very large range of "frequencies". Just set a CRT monitor to 60Hz to drive yourself crazy. There is no set limit to what the human eye can detect in terms of frequency, it varies from one person to the next. FPS is, in a twisted sense, a frequency. There is a "upper" limit that beyond would serve no benefit to the human eye.

The rule will always hold true, that the FPS one achieves or perceives will only be as good as the slowest calculation/frequency in the system or software. I.E. The bottleneck. You can only be as good as your weakest link.

60 FPS on a level 1 certified flight simulator makes sense. The most noticeable issue would be to ensure complete hand/eye coordination during maneuvers where the outside scenery being depicted is moving very rapidly. When flying FSX at 20,000 feet on a straight and level heading, at 30 FPS the simulation environment is smooth as glass. Get closer to the ground and stutters start becoming detected. Do a 45 degree bank around KLAX and watch what happens to the FPS, it plummets. If this was to occur on a level 1 flight simulator, this would be unacceptable. Keeping in mind of course, that the computers to run them are the price of a small house mortgage and use chilling units to keep them cool. Can we as normal desktop users ever really expect to achieve this kind of performance for the demands that we make on the simulation environment?

Hardware gets better, and then we demand more from it. It seems over the vast generations of flight simulators that we have always hovered around the same FPS mark, with exception to the earliest which had terrible lag. I believe if we can keep it around the 30-35 FPS mark, we will be doing good. I'd be very surprised if we ever break through it. 60 FPS, would be a dream, but if those kind of FPS were doable, creator's such as yourself would find ways to use 20 of them up to provide even more stunning visuals because you know the end user will accept the FPS hit, and would be happy if they could maintain the 30-35 range. My current build cost over $4000, it will run FS8 like silk with everything cranked on the sliders and still have time for a beer. On FS9 I could keep my FPS in the upper 40's, but am I prepared to use an older outdated less capable sim to get it? My answer is no. People want the eye candy but are not prepared for the consequences of it. They keep looking for a platform that will do the impossible, give them stunning visuals and frame rates in the stratosphere. Many look at X-Plane10, MS Flight, and Prepar3D with this sole focus in the forefront, and I see it as delusional.

If we want better frame rates from the current flight sim offerings, bigger changes need to be made. I.E. Get off the 32bit treadmill, utilize multiple core processors, and capitalize on multiple GPU's when present. Currently Lockheed Martin's Prepar3D team are the only ones that have even uttered the words of moving toward a 64bit solution and DirectX 11, something I see nothing of from it's counterparts. Maybe MS Flight will deliver, but seriously, has MSFS's ever met release expectations? If tighter financial times hit Microsoft again, do they once again punt the whole design and support teams, leaving all the users of MS Flight with a buggy program for the community to work around? X-Plane was designed with the Macintosh crowd in mind from the get go, will it ever be fully capable of utilizing the latest PC hardware, or be bound by the limits of current offerings by Apple? Is it's roadmap as a flight simulator in line with customer expectations? X-Plane has been around long enough for me to say for me personally, no. The progress of X-Plane to me is akin to watching continental drift take place. It is happening, but one has to look over a long period of time to see it. It still has limitations that Microsoft solved many versions ago.

My main point goes back to, there's more to any next generation platform than merely, does it give me more FPS. Logic would stand to reason as you so eloquently put it, that every new or improved platform will likely see gains in performance provided no new features are added to use the gains up, example FS9 to FSX. What is acceptable frame rates is a personal choice but I feel anything less than 30 should be evaluated carefully, folks should be wary of the "FPS" trap that so many seem to fall into in judgement of performance or wether a particular platform is good or bad or worth looking into. There are notable differences between FSX and Prepar3D in terms of performance, but it does not show on the FPS counter.

[Kilstorm]

Great news!  I've tried out P3D and feel in time I will replace my FSX with it.  This is based mainly on more addon developers like your FSDT supporting it with releases or installers that are P3D native.  So thank you FSDT supporting P3D!

[Highiron]

Any news on installers for the other products Umberto?

The KLAX version was out fairly quickly so I was wondering why the delay on the other airports. I have Hawaiian airports 1 & 2, KDFW, PHNL, and KLAS, and thought with the new Couatl engine in P3D I could just move the folders over from FSX and activate them in P3D. This was not the case. It classify's them active but not installed, even though the texture and scenery folders have been moved into the FSdreamteam folder. I suppose I'll have to contact Flightbeam Studios in regard to KSFO, even thought it uses the same Couatl engine for activation?

I would be getting KORD, KJFK, and the others, but since I've moved off of FSX, there's no point until I can use them.

[virtuali]

Any news on installers for the other products Umberto?

They are all up by now, I believe the only installer not updated to support P3D it's Geneva, because we have a patch in the works for it, but everything else is update, we just haven't fully announced it, because wanted to hear some feedback from P3D users how KLAX worked for them...

[Highiron]

Thanks Umberto, I've been downloading them all, except Geneva, I'll wait for the update.

Only it seems KFLL Fort Lauderdale's installer hasn't been updated. There is no prompt for P3D and it installed only to FSX.

Not sure about KLAS Las Vegas, as I'm getting a blocked URL as a malicious website.