Why do texture settings above 1m sharpen 1m textures?

[Guest Nick_N]

Actually yes, I can explain it and keep it simple tooIf you were to put my system side by side with yours and look at the monitors with both running the same settings

[Guest]

Yeah, there really is a difference! Although not every tree is smaller: some are bigger. It's mostly different, which is still weird because I thought autogen always was the same...

[Sesquashtoo 2,846]

Thanks for the post, Nick. Got it.BTW, what is TBM and what MESH would you recommend?Mitch

[Guest Nick_N]

TBM = Texture_Bandwidth_Multiplier inthe FSX.cfg fileMesh = FSGenesis 10m mesh and the World mesh that goes with itNote: If you are not a real stickler for details.. such as mountain range peaks and ridges in the right locations and valleys/gullys having a bit more realism to their real world counterparts.. install mesh is not going to make your eyes pop out..Its a wireframe the textures land on and better mesh means better match to real world but its not going to make the sim run better and unless you are obserbvant, its not going to make you jump up and down... its a realism factor, not a high end graphics upgrade.here is a post about landclass, mesh and textureshttp://www.simforums.com/forums/forum_posts.asp?TID=25569UTX is a landclass upgrade but it does not cover all areas. I use several other payware landclass products WITH UTX to put the finishing touches on things that UTX does not cover.. FSGenesis USASceneryTech North Americaetc.. Those are installed and placed in the scenery.cfg with UTX (correctly) and I only enable ONE at a time based on the area I am in. They will provide the data to GEXn textures to add things to the sim UTX does not cover.. like small one horse towns in the middle of nowhere, etc

[LuisFelizTirado 33]

Hello Geof,Strange about the autogen, perhaps you could take screenshots and document this behavior. It seems unlikely that autogen would change like that. Instead, it is more probable that the autogen varies somewhat and that the exact same objects are not always displayed in the same place.As for the small variation that you noticed, perhaps we could attempt to clearly understand what is happening.First, though, let me say that just like everybody else here, I do not know how Flight Simulator works. Like everybody else, my statements are pure conjecture, assumptions. And like everybody else, I like to think that they are logical and reflect basic common sense and therefore come close to the truth. However, this does not mean that they are correct, and I would urge everyone to come to their own conclusions and not blindly believe my claims or those of anybody else.Having said that, what exactly does the texture resolution slider do? As we can see from the above comparisons, there is no interpolation of lower-resolution ground textures to higher resolution at all. Any differences that people might see are purely due to display artifacts and the default ground textures display exactly the same at a setting of 15 cm. as at 1 meter.So, the slider simply seems to allow the selection of the maximum resolution MIP map to display. As I have shown various times before, all ground textures, in fact all textures in FS contain various bitmaps with different resolutions, for example:http://forums.avsim.net/user_files/186379.jpgBy setting the slider to 1 meter, the maximum 1024 x 1024 pixel image will be displayed. If the slider is set to 2 meters, then the maximum resolution displayed in game will be the 512 x 512 pixel bitmap, and so on.What does this mean? Setting the slider to a higher resolution than 1 meter (for example, 30 cm. or 7 cm.) will not display higher resolution ground textures. Why? Because the texture itself only contains a highest resolution of 1 meter, and that is the most that can be displayed.This slider then is mainly useful for custom (photo-real) ground textures such as the Edwards AFB ground that contains a maximum resolution of 60 cm. per pixel. (All the other photo-real ground textures that ship with the game have a maximum resolution of 1 meter, by the way.)Also, the slider serves to display higher resolution road and railroad textures. I assume that this is an oversight on the part of the guys in Redmond, as it seems incongruous that the roads do not display at the same resolution as the ground textures. Perhaps this issue is just due to the development rush.Anyway, here is one of the road texture with some of its MIP maps:http://forums.avsim.net/user_files/186381.jpgSetting the texture slider higher will display higher resolution bitmaps of the texture, and vice-versa.So, why do you not see a great difference in performance when setting the slider higher? Possibly there are a few reasons: - your brand new, extremely powerful computer can handle the larger textures without any problem. (The situation with my 5-year old computer is altogether different!) - these are not very large textures on the whole, nor are there very many of them, unlike the ground textures, so they are less demanding; - most importantly, you have not really placed yourself in a flight where you can adequately test this since the countryside in Michigan does not have a very dense amount of roads.I am not an expert on testing of code, again just like everybody else here, and that is why I consider any claims made in this forum to be essentially incidental and purely anecdotal.But, let us try to reason through the problem and find a situation that will give good test results.First, the flight must have a lot of roads, enough to stress the hardware when higher resolution road textures are displayed. This ensures that the results for different slider settings will show meaningful differences.Secondly, the test must isolate the element tested, the roads, and not allow any other factor, such as variable weather or traffic to affect the performance.So, I would suggest that if anybody is really interested in testing the effect of the texture slider on performance, they should do the following:1. choose a location with a very high density of roads, such as New York City or Los Angeles;2. slew to one thousand feet above the roads (probably not much more in order to ensure that lower resolution textures are not displayed because of greater distance.)3. remove the panel completely, only keep the forward view, do not on any account load VC, or any of the outside views;4. face North (press Control + Spacebar)5. Save the flight6. End the flight, in hopes that this will flush the cache7. Turn off anything that can affect performance, for example, all weather, also turn off weather variations, all shadows, bloom, water effects, all traffic of any sort8. Now, set the texture slider to a lower resolution, for example, 5 meter9. Load the flight and monitor frame rates (by the way, you do watch average fps and volatility, don

[Geofa 108]

Hi Luis-I don't doubt anything you say. I am just wondering though if autogen differences/ Nick's theory of devoting resources differently could be the cause. I did post two shots above showing the difference in autogen.http://www.mediafire.com/imgbnc.php/1b5baf...b9f427f694g.jpgMy blog:http://geofageofa.spaces.live.com/

[Guest Nick_N]

The write-up Luis provided is excellentits not that you are getting better texture resolution (except for streets) .. its that you are getting better priority placed on that part of FSX. Luis is absolutlely right.. you can not get better resolution than the texture installed. .. With settings changes is more time devoted to keeping things sharp and at the same time a side effect (depending on the system and how else it is set up) is FSX may be either skipping or removing other elements to provide that priority(rob Peter to pay Paul)the extra then goes to smooth flight instead of being directed at the activity the lower texture and mesh resolution allowedSaw it all the time with FS9. And I can put 2 systems side by side and have them show completely different views and responses on equal setttings, just different hardware.

[Geofa 108]

But my question was not about this-but the difference in autogen as submitted in the two photos above. The autogen on the higher res texture appears to be smaller.Seems to me smaller autogen might = faster fps and more realistic size?http://www.mediafire.com/imgbnc.php/1b5baf...b9f427f694g.jpgMy blog:http://geofageofa.spaces.live.com/

[Guest Nick_N]

oppsss sorry sir!I was still on the topic of the threadIn your case, smaller AG?Hmmmmm I can tell by your shots the system is not rendering a lot of AG and the textures have an image of the AG on them.. could you be seeing the fact that the system is not keeping up or is not set up to display a lot of AG and because of that it may appear to you to be smaller?When you raise the tex res it does set the mesh res to 10m automatically and therefore it will pull priority from elsewhere to make that happen.. so AG may end up reduced in rendering but I cant see it reduced in scale.Or if there are smaller AG components and larger ones... perhaps the priority is going to render as much as it can, meaning the smaller houses and not the larger onesNot sure.. but I know the slider is not going to change the scale of the same AG buildingEDITunless there is something I am not aware of about FSX and the AG library has smaller AG definition scale components for higher res texture and mesh calls.. never tried it before, interesting experiment to look into.

[Geofa 108]

Nope-I just tested again. It seems to be a part of the ftx package-the autogen appears to display differently whether running at 1 meter or 7 cm on the textures. The difference is very slight-but you can see in the shots above.http://www.mediafire.com/imgbnc.php/1b5baf...b9f427f694g.jpgMy blog:http://geofageofa.spaces.live.com/

[Guest Nick_N]

I will try it here with default and GEX when I get a chance later this evening and post what I find.Interesting if it happens to be the case. If it is true it may also explain a resource increase too. I have never noticed it on my system so it will be something fun to check

[Geofa 108]

I just tried that. I don't get the effect with gex, at least that I can tell. It will be interesting to see if others see the effect in ftx. I had not been running at 7cm till I bought ftx so this is a surprise to me.http://www.mediafire.com/imgbnc.php/1b5baf...b9f427f694g.jpgMy blog:http://geofageofa.spaces.live.com/

[Guest Nick_N]

Then it could be something specific to FTX designHolger could probably ring in on that one with information.Still, going back to the original topic, I find textures and scenery remain sharper with better performance @ 7cm and 10m anywhere in the sim regardless of add-on

[tf51d 284]

>I just tried that. I don't get the effect with gex, at least>that I can tell. It will be interesting to see if others see>the effect in ftx. I had not been running at 7cm till I bought>ftx so this is a surprise to me.>I'm seeing the same thing in FTX, no effect with GEX. but only with tree autogen. At 7cm the trees get noticeably larger, then at 1m, but buildings look the same in either.

[LuisFelizTirado 33]

No, Geof, not at all.Of the approximately 8000 ground textures, the majority displays autogen. Most of those have autogen annotations. An annotation is an indication on each of those textures of what should display there:http://forums.avsim.net/user_files/186411.jpgIn the image, every colored square was drawn in place (by me!) The thousands of default ground textures were annotated by the ACES Art Team and must have been a tedious drudgery of a task.For each building, the annotation indicates the type of building, the type of roof, and even the relative amount of floors per building for that texture.For vegetation, the annotation does not indicate a specific and unique tree, but rather a group of vegetation types. For example, although it is difficult to tell without consulting the original ground texture, from your images, it seems as if that texture displays one of the broadleaf vegetation groupings:http://forums.avsim.net/user_files/186412.jpgI have expanded one of those groups and you can see that it contains quite a number of trees, everything from Aspens to Sycamores. And each of those tree types contains different height variations as you can see when it is expanded:http://forums.avsim.net/user_files/186413.jpgFor example, you can see that the Elm contains different size trees from 8 to 17 meters.Any of the trees in a group can display on that annotation. Autogen display will change in different flights and there is never the exact same object in the same place. Nobody notices because, who looks at autogen? (besides me.)That is why you see different sizes of trees and not because there is any effect of a particular slider or mesh setting.Anybody who wants further information can consult an autogen guide that I put together and that contains an illustrated guide to the majority of the building and vegetation types.Best regards.Luis