Why can't t FS autogen look like FlyTampa's buildings?

[Guest JIMJAM]

I made a observation in the main post that we have addons replacing everything but scenery.Water,sky,clouds,grass,light,runways ect ect.Aerosoft's Manhatten showed that you can have tons of detail and buildings and still have a resonable frame rate.I am sure someone could replace the current autogen,scenery with a more performance freindly type.When FSX first came out, there was a effort to reduce all the scenery texture sizes but the performance payoff was small if any.Developers- We have enough clouds,water and grass. We want cities we can actually fly into without it turning into a slide show or turning down the detail til it looks like the surface of the moon.

[Guest Nick_N]

This is what GEX doesWe reorder and call autogen based on a strategic layout pattern. In doing so we provide a good visual fill and at the same time reduce the AG calls to the buss which is why GEX increases performance. We dont create new AG just reorganize the use of the default

[phoenixphire 50]

I have GEX and have noticed that it still affects my frame-rates enough that I do not fly with autogen.

[phoenixphire 50]

Interesting, I read your comment and another in the other thread. I did not realize that is where that thread progressed. I have heard that ACES is planning on changing the graphics engine from the ground up and that part of what is holding them back is backward computability. They would have to make a clean break and develop something totally new which would put an end to 3rd party addons at least until they were able to build addons for the new sim. In some ways i welcome this. In the past it seems like every advancement in the sim meant one thing, better hardware. More detail in the sim equals more raw CPU and better graphics cards. FlyTampa has proved that you can add incredibly detailed scenery and not suffer from it. In the other thread George writes about the sims SDK being 10 years behind.If this is so then a rewrite might actually leave the sim in an entirely new situation. the software technology has caught up with the hardware and now with the proper rewrite we could put an end to the constant cycle of FS release followed by trying to catch our hardware up to the new sim. I know this is a very untechnical simplistic way of looking at the issue. Am I off base here or could this be a reality? Just wondering.

[jackcnd 0]

WOW, just looked at some flytampa Hong Kong FSX screenshots.The thing is out of this world.I thought GEX was pretty good but this just raised the bar 10 feet :)This thread shows default FSX and then flyTampa shots:http://www.flytampa.org/forum/viewtopic.ph...der=asc&start=0

[irocx 292]

>>>This is what GEX does>>We reorder and call autogen based on a strategic layout>pattern. In doing so we provide a good visual fill and at the>same time reduce the AG calls to the buss which is why GEX>increases performance. We dont create new AG just reorganize>the use of the defaultNick, I believe the question the OP posed was: "Why FS autogen can't look like FlyTampa's buildings?" Not about GEX or what it can or can't do.Not a huge deal Nick but, I continually read "we" in your various posts when referring to GEX and yet I see no affiliation in your signature. Your post appears like a blatant sales pitch for GEX, at Fly Tampa's expense no less - not to mention it has little to do with the OP's original inquiry. And yes, I do see all of the excellent posts you've made assisting fellow simmers - kudos to you and thanks. This post just seems very blatant to me though and I repectfully believe it's time that you make your affiliations obvious to everyone. JMHO.Regards, Kendall#1: E6750@3.2GHZ/Coolermaster HyperTX2 Gigabyte P35-DS3L 4GB Ballistix Tracers PC6400 EVGA 8800GT - 174.74 beta Seagate 250GB 7200.11 CH Yoke/Pedals/Saitek Throttle Dual Monitor: Dell 2405/1905 #2: Dell 8400 3.2 H.T. 3GB PC4200 - X800XT Diamond Xtreme/Logitech X-530's

[Guest Binkles]

Not just for FSX, but it looks bloody awesome in FS9 - I can't believe that this is the first time I've seen a fully developed area to this degree.I think all other scenery designers have a long, long way to go to catch up to this.What's amazing is that it's for both fs9 and fsx - good to see someone still working on fs9. Shows that a rebuilding is not definitely required

[LuisFelizTirado 33]

The more scenery detail that is added, the greater the impact on performance. This is an absolute truth and there is no way around it.Martin is very honest about his work and in his forum he clearly indicates the following:Like other large urban airports in FSX, you should not expect being able to fly a complex aircraft like the PMDG 747 into Hong Kong with Maxed AI-Traffic, Maxed Display-Settings and complex Weather Themes the same way you fly a GA aircraft over wilderness scenery. Before you purchase, it is recommended that you take a look at FSX's default Hong Kong (VHHX). If you can't achieve a minumum of 30 FPS at the default Hong Kong, your computer is not suited to run this Addon in FSX. Any detailed reproduction of urban areas will add so many objects (individual buildings and trees) that performance will suffer heavily. Autogen diminishes the performance hit by repeating instances of the same object - they are not unique and only must be rendered once for each. Still, too many objects will always affect performance.Anybody who does not see an impact has a computer powerful enough to handle the load. But, everybody else with less powerful systems will see signigicant impact.Best regards.Luis

[michal 44]

>FlyTampa has>proved that you can add incredibly detailed scenery and not>suffer from it.I am sorry but this is a complete fallacy. It is like believing in the perpetuum mobile. You can't even compare world-wide autogen with a limited area autogen like in this case. If your computer can handle and you don't see much impact - consider yourself lucky. Michael J.http://img142.imageshack.us/img142/9320/apollo17vf7.jpg

[Guest jshyluk]

That's not to say that there can't be improvements made in FSX scenery. FSX SP2 is a good example of that.The idea runs like this: you make a building that has an odd number of polygons, say 43. However, you can make the same building with 42 polys, because 42 is a much nicer number. To the untrained eye, the two buildings are identical. The trouble arises when you draw that one 43-poly building a thousand times. You now have 1,000 polys that you don't need. If you can tell the renderer to expect all of the buildings to have 42 polys, and it does not have to search for buildings that have 43, that's even better. From what I know, that's what SP2 does, although my numbers are purely made up. The advantage is that the buildings are more standardized, and that the renderer does not waste cycles on so many non-standard buildings. That helps reduce the load on the draw cycles overall, and should contribute to frame rates.Polys are just one thing that helps frame rates. Lots of things help the frame rates, but mostly in combination. I can't really think of any one thing (except maybe jetways or traffic) that if you got rid of it, you'd see a big jump in framerates. So to get better frame rates from models, you have to have all of your art teams working from exactly the same standard. That being said, Luis is right: there's going to be an upper limit to what your computer can do, and once you hit that limit, tweaks just won't help.Jeff ShylukAssistant Managing EditorSenior Staff ReviewerAVSIM

[Guest Nick_N]

>I continually read "we" in your various posts when referring to GEX >and yet I see no affiliation in your signature. Your post appears >like a blatant sales pitch for GEX, at Fly Tampa's expense no less -> not to mention it has little to do with the OP's original inquiry.That is a very strong statement, practically an accusation

[GHarrall 494]

Nick,Forget it. You didn't need to apologise for anything.This is why I spend most of my time over at Simforums these days......

[loop_n_roll 7]

Nick,Most of us here recognize your unselfish contributions to our hobby as well as your contributions to UT and GEX. I eagerly read each of your posts. No need to apologize and those of us who know you realize you were enlightening us without criticizing anyone. Thanks.

[bob.bernstein 0]

Nick, I didn't think there was anything wrong with your post. Apology seems un-needed.In fact, my reading of the op's message left me with a thought that someone should correct the discussion to eliminate the comparison of the "top of the line" scenery offering to the "general" scenery expectation. Frequently I bristle when people get excited over "top of the line" designers work, and then say..."why can't everyone be this good". Well its becuase skill is developed on a continuum. The longer you do it, the better you get. Fly Tampa will get even better as long as they continue practicing, so even their level of quality is a moving target. Everyone else is where they are, designing with the skills they've developed at the moment, enjoying the movtivation of role models like Fly Tampa, and serving as role models for designers still earlier in their development. I've found that my skills at the end of a project are evolved from my skills at the beginning of the project, so the only way to ever release is to release some content below my own knowledge level. Its just a continuum, and release of any project is a snapshot of skill at that point.Aces have had the same experience. Look at Istanbul with full autogen (with acceleration/sp2) and compare performance with Seattle. Improvements have been learned. Beyond the notion of continuum is the notion of design goals that the OP overlooks. Does the OP think that MS intended autogen objects to appears as focals for viewing or as elements that together form the sense of expanded 3d to the experience. Clearly the latter, yet Fly Tampa expects entirely different use of their objects, their objects are art to be viewed carefully by the customer, and not to be found willy nilly all over the world. You effectively righted the ship of logic, by pointing out improvements that are available today through the use of current objects in a way that vastly improves their impact on immersion. GEX is masterful, one of my favorite addons, in that it takes the objects intended for fleshing out the 3d experience, and orients them to ground imagry to expand the immersion.The one idea that partially explains the OP post and the latter objection is if the comparison of Fly Tampa to MS autogen was just a smokescreen to hide the real intention of the post, which was just to praise Fly Tampa. I'm good with praising Fly Tampa, but If I want to do that, I leave out the wacky comparisons and just simply praise them. The wacky comparisons are why I thought this was NOT a thread intended to praise Fly Tampa, except as a passing thought.Perhaps the OP thought simply praising someone wasn't "meaty" enough to merit a post, so he tried to enlarge the praise into something bigger? Cheers,Bob

[FlyingsCool 3]

"The more scenery detail that is added, the greater the impact on performance. This is an absolute truth and there is no way around it."Yes and no.One of the things you can do is improve the textures and theoretically this may not have any impact on frame rates. No knock on ACES, they have a huge job to do in a small amount of time, and then we get years to improve on their work. But I've seen some amazing artistry out there, taking a flat texture and, just by modifying pixels, giving it a 3D look with better shadows, etc., etc.. (which is really tough to create given all the technical challenges which I won't go into here).It's been done before.I think the original point of the thread was that - hey, there's people out there modifying the ground textures; how come nobody's gone after the a'gen textures yet?I also believe that ACES opened up more slots as well for different types of a'gen. While it is a given that adding more types of a'gen will probably affect performance, I think there is room for improvement as well.Of course, there's also the issue of life expectancy. If things really improve in FS11, I'm imagining that FSX will be dropped by users a lot more quickly than FS9 has been. Given that it would probably take 1 or more years to develop such a product, there's probably not much point now.Thomas[a href=http://www.flyingscool.com] http://www.flyingscool.com/images/Signature.jpg [/a]I like using VC's :-)N15802 KASH '73 Piper Cherokee Challenger 180