>>Aaaah,>>>>Still looks pretty. Did you have the chance to look into the>>mirrored roundel thingy I posted in the other thread. (the>low>>vis SF 360's right wing). This one at least has the right>side>>up.>>>I guess I don't remember the other thread.>>Glen Canyon, happily only an hour and half away by air! >Its about this picture you postedhttp://forums.avsim.net/user_files/171896.jpgThe right wing national markings are reversed ;)

No, XPlane's textures look good at all altitudes at any speed, they just don't swap in & out. FSX textures can look good too, in a freeze frame at optimal altitude.Not gonna second guess how each product is coded. Nor am I saying Aces scrap everything -- is removing texture swapping a huge deal? It certainly would permit more CPU cycles to work on other tasks of the sim if it weren't constantly switching texture maps. But something needs to be done about LOD and textures cause the current scheme isn't good regardless of your level of tolerance. It really isn't taking advantage of the video card VRAM available today, sure it might be the best way for someone with 128MB -- but if someone has low end hardware I would expect that person to be VERY tolerant.Nothing wrong with making a product scalable, Aces have started in that direction with multiple CPU support, why not continue on in that direction of flexibility and scalability so that as hardware does improve so does FSX. Phil Taylor did mention that they weren't doing anything with VRAM >256MB til SP1. Why place artificial limits on allocation?Try not to exclude the lowest hardware, but at the same time don't impose a penalty (meaning don't use a design/method that works well for low end but ignores potential benefits of high end) for those that do have the hardware. Make it scalable.But the bottom line is do whatever it takes to get rid of the texture swapping scheme so that blurries (and/or shape shifting) can never be mentioned again. Who does or does not have tolerance for the blurries really is NOT the issue -- they exist on more than just a few systems.

>No, XPlane's textures look good at all altitudes at any>speed, they just don't swap in & out.You must have stated your opinion about this "bad" texture swapping hundreds of times. And you keep presenting instead XPlane's scenery as an example of 'virtue'. Yet we know that Xplane's limited radius of 25 miles is VERY limited which allows avoiding some rather nasty scenery design decisions. We know that terrain grows very quickly as radius grows - as fast as square function so situation worsens very quckly. Lets see what happens if XPlane lifts the 25 nm restriction and whether we get blurries, stutters, etc. At least we will be comparing apples to apples and not apples to oranges. In my humble opinion even with occasional blurries or stutters FS's scenery looks so much better - all its elements come together and fit together. XPlane's scenery unfortunately looks good when you examine its separate elements in isolation (mountains, buildings, etc) but the final look-and-feel is very unconvincing. My opinion of course. If and only if XPlane's scenery can reach sophistication of MSFS's with all its radius potential then we will be able to judge which scenery engine is superior.Michael J.http://img142.imageshack.us/img142/9320/apollo17vf7.jpg

At this moment in history there is 1 flightsim series that gives us unlimited visability. Pro Pilot, Flight Unlimited, Fly, X-plane they all have hard limits to the maximum possible visability. Is having unlimited visability in the FS series realy an asset?As you can see from my previous post. If I can trade-in visibility (my self imposed 33 - 40 km max I like to set) and in return I get a a fluid 30 fps and no blurries then I need to think about 10 seconds before I say YES.Most FS use I know of can be perfectly done in a visability limited scenery engine. The saved performance can be used elswhere to improve the flying experience. I get the idea that FS-X is doing way to many thingss that in the end don't matter for my experience in the game itself.

>Is having unlimited visability in the FS series>realy an asset?No, and I never stated that. But there is a 'sea' of difference between 25 miles and say 150 miles. It is highly unrealistic to fly above California (summer or winter) at 30,000 feet and be restricted to 25 miles. It is simply poor simulation of basic atmospherics.Look, if someone is happy with limited radius of Xplane's scenery and thinks it is a worthy trade-off for performance - more power to him but do not use it to flout MSFS's scenery - it is ridiculous.Michael J.http://img142.imageshack.us/img142/9320/apollo17vf7.jpg

>At this moment in history there is 1 flightsim series that>gives us unlimited visability. Pro Pilot, Flight Unlimited,>Fly, X-plane they all have hard limits to the maximum possible>visability. Is having unlimited visability in the FS series>realy an asset?>You had better believe it's an asset! I can usually see mountains up to 80 miles from my back door, and well over 120 or more miles when airborne.I always had problems with those simulations you've mentioned, when I could never see both sides of the Salt Lake Valley at once. It was as phoney as can be, and unconvincing. L.Adamsonedit: after some more meaningful flying around with X-Plane, global scenery, and the Salt Lake Valley; X-Plane does a better job of getting both mountain ranges than in the past. It does prompt me to lower my viz settings because fps want to drop into the low twenties.Scenery wise it's all still compromises though, whether it's FSX, FS9, or X-Plane. Just depends where, time of day, and complexity of scenery.

I agree with Boshar, I'd take a 25mi max any day over shape shifting 4.5+ mi radius (which I believe is what the max LOD value is for FSX) where one has to wait 2-7 seconds for the scenery to morph everytime a view is changed.But rather than presume FSX scenery engine is superior, why not come up with scalable alternatives? There are many ways to approach a particular problem, with current video hardware this problem needs to be re-worked so that it can meet the potential of current hardware and future hardware.As resolutions continue to increase the texture loading issue is only going to get worse not better. There is no miracle hard drive or caching system now or in the future that will help -- ultimately the final destination will be textures ready to roll in VRAM.Take a look at nVidia's road map to their next gen cards, more VRAM, more VRAM, more VRAM. I assume ATI will be doing the same. This is where you want your textures.I really don't think this would require any "massive" change in FSX scenerey engine -- certainly far less difficult that dealing the threading on multiple CPUs. Sure, it may require some changes in how scenery is loaded but ultimately the scenery data is there.1920 x 1200 32bit = 8.2MB per pixel buffer, pending what processing one does there would be 3 or 4 buffers allocated at any moment in time (about 25MB) leaving a rather significant chunk for texture cache -- the hardware programmable renders might consume and what compression is or isn't used, etc. etc. Unless I missed something, there should be PLENTY of VRAM available for texture caching on today's 512MB on up to 2GB capable VRAM graphics cards.Microsoft's tools help with profiling: http://msdn2.microsoft.com/en-us/library/aa969767.aspxscroll down to PhotoDemo.vshost in Perforator where est. video memory usage -- I know FSX uses VS C++ 2005 but not sure if WPF can be used - haven't tried it yet.But either way, there should be plenty of VRAM available for texture caching even with a 50mi LOD.

I agree with you, but which is more unrealistic? Shape shifting scenery beyond 4.5 mi radius or a fixed 25mi radius?I'd rather see atmospheric fake haze than an ever morphing volcanic activity and CPU hogging scenery even at 30,000 feet.

>I agree with Boshar, I'd take a 25mi max any day over shape>shifting 4.5+ mi radius (which I believe is what the max LOD>value is for FSX) where one has to wait 2-7 seconds for the>scenery to morph everytime a view is changed.>For starters, 25 miles would be totally unacceptable. I live, fly, and sim fly in the mountain west, where clear air far exceeds those limits. To have limits such as that, for a bit of "shape shifting" would be rediculous for the types of flights I prefer.Secondly, it was good old FS2000 that took many irritating seconds to morph. FS9 and FSX is really a non event as far as morphing is concerned. BTW, X-Plane just plain stops for a solid 10-15 seconds for scenery reloads. Might be 10-20 minutes in between though.L.Adamson

>FS9 and FSX is really a non event as far as>morphing is concerned.I agree. Morphing has ceased to be an issue for me too some time ago.Michael J.

This whole thread is why I think Tileproxy generated scenery is the future of flight sim. I don't find the generic fsx textures, clumped by landclass that is rarely correct, with autogen houses and trees popping up that look like sim city 2000, and roads that look like lines drawn on the scenery very realistic anymore. It was the best we could have for many years and I was appreciative to have that type of reality in the past-but the transition time is now. I think Aces must have anticipated this by allowing fsx to display the high res photo textures. As on the Tileproxy thread-there is already a technology to convert 2d photos into 3d-it will only be time when the photo real scenery sprouts real buildings and trees where they actually belong.After experiencing Tileproxy I can't go back to the generics-why when I can get stellar performance on scenery that finally duplicates the real world tree by tree, house by house, road by road, man made and natural formations-today? No stuttering, no visibilily issues, fps almost twice as high as the vector driven computational hog that the present scenery requires. There are still issues to work out with Tileproxy type technology (blurries of a different kind, and autogen) but I find the limitations of Tileproxy much less even now than the limitations of poor performance, blurries, incorrect landclass, and popping autogen.http://mywebpages.comcast.net/geofa/pages/rxp-pilot.jpg

> I don't find the generic fsx>textures, clumped by landclass that is rarely correct, with>autogen houses and trees popping up that look like sim city>2000, and roads that look like lines drawn on the scenery very>realistic anymore.Hmmm... I guess I'm kinda hooked on autogen, though there's certainly room for plenty of improvement in how it looks, how it is added (popping) and how it interfaces with 3rd party photo textures. Photo-real is great once you get up there, but a pancake world really bugs me when I'm on final. Hopefully the DX10 patch will mean another good step forware in autogen appearance... though I would guess being able to interface with 3rd party photo-real is a little bit off.

Probably because one is DX based, and the other is OpenGL based. It's a known fact OpenGL is optimized much more than DX. I myself haven't had any trouble running OpenGL based applications vs. running DX. DX is what's always given me the problems. I don't much care if I get flamed for this post or not, but I'm a die-hard when it comes to OpenGL and will stand by it. Anyone hear of an ESIG, etc., running DX? I've yet to see it.

>Probably because one is DX based, and the other is OpenGL>based. It's a known fact OpenGL is optimized much more than>DX. I myself haven't had any trouble running OpenGL based>applications vs. running DX. DX is what's always given me the>problems. I don't much care if I get flamed for this post or>not, but I'm a die-hard when it comes to OpenGL and will stand>by it. Anyone hear of an ESIG, etc., running DX? I've yet to>see it.Perhaps you have valid points/beliefs about OpenGL, but I think this interjection is misplaced. I'm fairly certain that has little to do with this debate. You could implement FSX's LOD terrain system with OpenGL and likewise you could implmenet a static terrain system with DX.Anyway, game developers outside of M$ have every opportunity to use OpenGL instead, but few do. I'm sure OpenGL has its advantages (and vice-versa), but if OpenGL was a magic wand then why the predominance of DX?

>Hmmm... I guess I'm kinda hooked on autogen, though there's>certainly room for plenty of improvement in how it looks, how>it is added (popping) and how it interfaces with 3rd party>photo textures. Photo-real is great once you get up there,>but a pancake world really bugs me when I'm on final. >I see that Virtual Earth is already getting populated with 3D buildings for the KSLC area, which is where I fly from, not to mention the excellent looking 3D mountain/foothills. And Las Vegas is starting to look real good with lots of buildings, swimming pools etc.! I find that with a bit of altitude, even photo trees take on a 3D effect, depth wise. Of course some images are five years old or so. But I'm hooked on "high resolution" photo-real, or sensational looking scenery like FlightScenery's FS9 Portland. Still haven't gotten around to installing tile-proxy. IMO, some good 3D airport buildings while coming in on final will work just fine, for now!I must say that my favorite landings are at some of those FS9 "Flight Scenery" Portland area aiports, with crisp & focused 3D scenery along with moving vehicles as I pass over the country looking fence! :)L.Adamson