Tuesday, April 01, 2008

Instaling to the Desktop (Let's Start a Riot)

I am intentionally announcing a decision that might make experience X-Plane users unhappy - my goal here is to solicit arguments against it (if there are any) since it's a sort of a strange change.

The proposal is to make the default installation location for a new copy of X-Plane be...the desktop.

EDIT: to clarify some of the blog comments, this is a default installation location; the user will be able to customize both the folder name that is created to contain X-Plane (default will be "X-Plane 9") and the folder into which this new X-Plane 9 folder will be placed.

The desktop? What are you hacks thinking? Well, here's what we're thinking:
  1. Our goal is to minimize tech support calls during installation. This means making an installer where the last computer-savvy users will not get stuck in the installation process. If you know what you're doing, you're not who we're aiming at. (I reiterate, we get a lot of calls about installation problems from users who have never used a computer before.)
  2. We need a location that the user installing X-Plane has access to, guaranteed. That rules out places like the Applications folder on OS X - we expect the least sophisticated users to not customize the install location, so we need one that will work.
  3. We need a location that the user can find. This rules out their home folder (the user may not use their home folder or know it exists) as well as the C drive and the Program Files folder on Windows (both can have their files hidden by default on Windows XP to keep users from breaking tings).
The desktop solves all of these problems. Every user has write-access to his or her own desktop, and every user knows where it is.

If, like me, you don't want X-Plane on your desktop, you can simply click the "destination" button in the installer to put it somewhere else.

As a final note, the strongest alternative to this on Windows was to put the app in program files and build a start menu short-cut. But this starts the sweater unraveling...if we have a shortcut, we need an uninstaller rather than trashing the folder...if we have an uninstaller, how do we cope with multiple installs...by the time you solve these problems you have a huge amount of new untested code.

I'd like to get a more Windows and Mac interface compliant installer, and we'll get there eventually, but the work I'm doing now is aimed at the biggest real problems we face:
  • Users not knowing where X-Plane is.
  • Errors during the install in its default configuration.
  • Problems installing and configuring scenery.
Okay, there it is, fire away!

Sunday, March 30, 2008

DataRefs, the Cirrus, and Opt-In

There's a bug in X-Plane 900 RC2 that will be fixed in RC3.  Basically when you're flying in the Cirrus with a friend (also using the Cirrus via multiplayer) you'll see your own control deflections on his or her plane.

Understanding this bug gets into the way datarefs and object animation work.  In simple terms, a dataref is an access point to data provided by some other part of the sim...the flight model, or another plugin.  An object references a dataref as its source for animation values.  So if you want to animate the wings, you use sim/flightmodel/controls/lail1def or sim/flightmodel2/wing/aileron1_deg.

Now the question is: when you access this dataref in a multiplayer context, which aircraft's ailerons are you talking about?  It turns out the answer depends.
  • For all datarefs except sim/flightmodel2/ and sim/cockpit2/, if the dataref needs to access a plane, the access is to the user's plane (flight 0), always.
  • For sim/flightmodel2/ and sim/cockpit2/, the access is usually to flight 0.  But if we are drawing an aircraft, these datarefs refer to the aircraft being drawn!
(Note: this second feature only works correctly in RC3.)

So when you want to animate the wings of your plane in X-Plane 9, by using the sim/flightmodel2/ datarefs in your object, you'll get the behavior you want: animation with the first plane's flightmodel when your plane is loaded as plane one, animation with the second plane's flightmodel when your plane is loaded as plane two, etc.

Of course the rule of thumb is simple: if working on v9, always prefer sim/flightmodel2/ and sim/cockpit2/ when possible for animation work.

Could we have simply made the sim/flightmodel/ and sim/cockpit/ datarefs have this "pick the right airplane" behavior?  Possibly, but I took the more conservative opt-in approach.  Basically any time we want to change the behavior of the sim we can do one of three things:
  1. Change it globally - all third party content is affected; don't like it, too bad.  This is how pixel shaders work; if shaders are on you can't disable the different fogging they provide for your airplane or object.
  2. Opt-in - you have to change your content to get the new behavior; old content defaults to unchanged.  This is how the new datarefs work.  You have to edit your content to use the new datarefs to get the new behavior.
  3. Opt-out - you have to change your content to avoid the new behavior.  For a while cockpit lighting was like this, but I received so many reports of version 8 airplanes looking totally strange that I changed the cockpit lighting choices to opt-in.
Generally opt-in is the safest thing we can do to leave previous content working; if we make a change globally then we break any content that doesn't work well with our changes.  Datarefs are used so heavily that a fundamental change in how they work would probably not be safe.

Given a choice between opt-in and opt-out I've come to prefer opt-in; both require the same amount of work for LR (we have to write code to support the old and new behavior) so why not do the thing that provides better compatibility?

Finally a general historical note: sim/cockpit2/ and sim/flightmodel2 are not just there to provide new behavior with objects; they are designed to give us a clean start in providing only authoring-related datarefs in an easier-to-understand format.  The original datarefs were meant only for programmers, so we didn't worry too much about usability.  Since then, datarefs have become the communications medium for panels and a lot of 3-d animation.

So not only do sim/cockpit2 and sim/flightmodel2/ provide correct animation behavior, but they're also easier to read.

An important distinction between sim/cockpit2 and sim/flightmodel2 is that they partition the datarefs based on perception vs. reality:
  • Generally sim/cockpit2 shows datarefs as the pilot sees them in the cockpit, and actions as the pilot commands them.
  • Generally sim/flightmodel2 shows the plane's behavior as it's really simulated, and actions as they actually happen.
In other words, when the pitot tube ices up, sim/flightmodel2 shows the real airspeed and sim/cockpit2 shows the incorrect airspeed.  When the hydraulic system fails, sim/cockpit2 shows that the pilot wants to bank, and sim/flightmodel2 shows that the ailerons aren't really moving much.

What this means is that you should use sim/cockpit2 for cockpit objects and generic instruments and sim/flightmodel2 for objects attached to the airplane exterior.  This will assure that failure simulation works correctly (e.g. failures are perceived correctly by the pilot) and the external control surfaces match the plane's physical behavior.

If you've studied sim/cockpit2 a lot, you know it's pretty incomplete.  We'll be adding the rest of the cockpit systems in 910.  I would have liked to get the entire set of airplane changes into 9.0 but there's a lot more to do - not just more datarefs, but also manipulators for 3-d objects and a new lighting model for aircraft interiors.  I think 910 will finish what 900 has started in terms of new airplane features.

Final random note: RC3 will allow you to pick any dataref for a generic instrument, not just sim/cockpit2 and sim/flightmodel2.  This is mainly so people writing plugins can easily hook their datarefs up to the panel by editing datarefs.txt instead of typing them by hand.

Wednesday, March 26, 2008

Road Overlays and Draping

X-Plane 9 allows you to put roads in an overlay DSF, but this isn't actually that useful yet.  The problem is that roads are specified using MSL elevations - that is, they are positioned in absolute space.  So you have to know where the ground is to build a road, which defeats the whole idea of an overlay.

This dates back to X-Plane 8.0 - originally all scenery load was done when a DSF was loaded, with the sim paused.  In that condition, keeping scenery load fast was a high priority.  So I set roads up to be "pre-draped" (that is, their heights are pre-determined when building the DSF). Back in the day it was a win.

Now we have multiple cores and DSF loads in the background, so we very easily could "drape" a DSF...but it gets weirder.  A road's sub-type is specified in the DSF - and sub-types are used to define bridges.

Now that's not incompatible with a "draped" scheme - imagine a road whose height is specified AGL, so a height of 0 means "run the road along the base mesh" and a height of 10 means "10 meters above the base mesh."  Now you would set the road type to overpass as the height went from 0 up to 10 meters, then back to 0, forming a bridge.

The only problem is what happens if the ground is really weird looking.  In practice what we want to do is keep the road's slope as constant as possible and build out bridges based on that. The effect can be hard to see, but if you look at bridges in the US scenery, you'll see that sometimes the banks to a river slope down but the road stays horizontal, forms a bridge, and then stays flat as the banks come up again.  This happens because our road processing code (in the DSF builder) tries to limit road slope.

That's not an easy effect to create in an overlay...you'd basically have to increase the road AGL at the rate the base mesh's AGL is decreasing (if only you could know that).

So the questions I am debating now are:
  • Are AGL-draped roads useful if we don't know what the base mesh will be like?
  • Does it make sense to have some other abstraction to help smooth roads when the base mesh is bumpy (and can we make something that is both simple and useful)?
Like most philosophical things I debate, this isn't scheduled to go into the sim that soon. Improved shaders, a more complete airplane SDK, and a better texture pager are all ahead of road improvements, which pushes roads out quite a bit I think.

Tuesday, March 25, 2008

You Can't Replace Objects in Liveries

The livery system lets you replace the image files used by an aircraft's exterior paint, and it lets you replace the image files used by the objects that are attached to the plane via the "objects" folder.

But the livery system does not let you replace the objects themselves.

(It also doesn't let you replace or modify the actual ACF file or generally change the functional behavior of the plane. The livery system is just for paint, not airplane mods.)

Friday, March 21, 2008

More Threads - the Installer

Nine women cannot have a baby in one month - that's the classic example that gets thrown around computer science for the difficulty of parallelization - that is, just because we have ten times as many resources doesn't mean we're going to go ten times as fast.

Problems of scalability via parallelization have become very important for graphics engines as everybody and their mother now has at least two cores, and users with more serious hardware are going to have four.

I get asked a lot: "will X-Plane utilize X cores..." where X is a number larger than two. My general answer is: sometimes, maybe, and probably more in the future. I can't make strong predictions for what we'll ship in the future, but the general trend for the last 18 months has been us using more cores every time we go into a piece of code to do major architectural changes.

I've been doing a lot of work on the installer this week - the first major overhaul of the installer since we originally coded it all the way back at X-Plane 8.15. And the new installers and updaters will try to take advantage of multiple CPUs where possible. A few cases:
  • The X-Plane updater runs an MD5 checksum over the entire X-Plane folder to determine which version of the various file components you have and whether they need to be updated. This s not a fast process. I am working on threading this so that more CPUs can work on the problem at once. It looks like there will be only modest benefits from this because the process is also highly bottlenecked on the disk drive.
  • The installation engine from the DVD will use more than one CPU to decompress files. For zip compression this wasn't very important, but the scenery will be compressed via 7-zip compression to get us down to disk DVDs. 7-Zip compresses DSFs about 10% smaller per file than zip, but it's horribly slow to decompress, so being able to throw twice the CPU at it is a big win.
Now on one hand, our top performance goals are for the sim, not the installer. On the other hand, faster installations are good. But my main point here is: when we wrote new code four years ago, we assumed one CPU and a nice graphics card. We now assume at least two cores and possibly more, and that informs the design of every new feature, not just the rendering engine. If we don't create a multi-core-friendly design, we're effectively leaving at least 50% of the CPU on the table.

Wednesday, March 19, 2008

When Will X-Plane Stop Saying It Is Beta?

If you look at the about box for X-Plane 9 RC1, it still lists itself as a beta.

Here's the thing: that label is dynamic. Right now when X-Plane calls up the server to see if there is a new patch, it notices that its current version (900Rc1) is newer than the latest "final version". (This is because we haven't declared any version of 9.xx final yet.)

So it thinks for a second and goes "oh - I'm newer than the latest final version, I must be a beta." And that yellow beta label appears.

When we finally declare a version final (which involves tweaking the versions listed on the servers) the existing code will then look at the server and go "oh, I'm the latest version" and that beta label will disappear.

Tuesday, March 18, 2008

The Global Scenery Folder

Let me answer two questions right now:
  1. If you bought version 9 beta DVDs, you will not have to buy new ones when version 9 goes final.
  2. If you bought version 9 beta DVDs, you do not need to manually move around any of the global scenery files that our installers put on your hard drive.
Now about this global scenery folder...

X-Plane stores its scenery in packages - a scenery package is a folder that contains all of the files needed for a given scenery element, whether custom or default. In X-Plane 7, scenery packages were only used for custom scenery, and always lived in the Custom Scenery folder directly next to the X-Plane application. This provided an easy way for users to drag third party packages into a single location to install them. (Before the custom scenery folder, installing scenery could involve a lot of "put this file here, and that file there".)

X-Plane 8 added a second location, the Default Scenery folder, which is hidden away inside the resources folder; it acts as a repository for scenery packs we ship with the sim, such as the one that contains various airport elements, and the default road packs.

In X-Plane 7 the global scenery lived directly in the resources folder in its own home; in X-Plane 8 the global scenery lives in packages inside the default scenery folder...this allows us to use one piece of code (the package loader) to load all scenery, custom and global.

The only real differences between the default scenery and custom scenery folders was priority and intended use; custom scenery is loaded with higher priority than default scenery (so you see your custom add-ons replace our default work), and the intention is that the default scenery is for Laminar to update via the web-updater and custom scenery is for you to put things into that you download off the web, etc.

With X-Plane 9 we are introducing a third location for scenery packages: the "Global Scenery" folder. It will live next to the Custom Scenery folder and X-plane application and be the location for global scenery. Its priority is intermediate - higher than default but lower than custom.

Now here's the funky thing: the X-Plane 9 beta DVDs place the global scenery into the custom scenery folder. If you've ever looked at the ranking of scenery packages, you'll see that they come out lower priority than all custom scenery, regardless of the alphabetical sorting that is usually used. The sim recognizes the peculiar placement of the global scenery and effectively labels it "global".

Future DVD pressings will simply place the global scenery in the new global scenery folder. Thus we get to the original two answers:
  1. You won't have to buy new DVDs if you have the betas; the sim will continue to work with either the beta or the newer layout for global scenery for the entire v9 run.
  2. Because both layouts work you don't need to move anything around. I suggest you not move any files to lower the risk of breaking things.

Thursday, March 13, 2008

Scenery Tools Release

I have posted the first "combined" scenery tools release. Basically ObjView, ObjConverter, and DSF2Text (now named DSFTool) are packaged together with DDSTool (for making DDS images from PNGs) and the new MeshTool (for making base meshes); a UI application called XGrinder provides drag & drop support for most of the converters.

The tools releases will be by month (e.g. March 2008) and each release will contain the latest for all of the tools in the release. WED and the AC3D plugin will continue to be separate downloads.

I have also posted another snapshot of the source code, and generally I'll try to do a source code post whenever I do a tools post to keep the two in sync.

The AC3D plugin has been updated; I hope to have a new WED posted in the next few days.

Tuesday, March 11, 2008

X-Plane 9: What Comes Next

X-Plane 9.00 is going to be going final pretty soon - we're on RCs right now. But this is hardly the end of the road; rather it's the first step for a number of new development areas. What comes next?

Tools: I am working on tools releases now; a few users already have alpha copies of MeshTool. My hope is to have some new tools posted tomorrow, with more by the end of the week

More rendering engine/shader work: X-Plane 9 only begins some of the shader work we want to do; more engine enhancements will be coming in 910. Like 9.00, more advanced rendering will consume more hardware, and will be scalable via rendering settings.

Systems and Airplane SDK: X-Plane 9 introduced a lot of new features for airplane modeling, but there's still more to do. I hope to have the panel region system and advanced lighting options all fleshed out for 9.10. Austin is doing a lot of work on systems modeling. We'll be making more new airplane-related datarefs to tie it all together.

Orthophoto Scenery: I don't know if this is going to happen or not, but we get more and more requests for large-orthophoto-scenery support in X-Plane; MeshTool is only going to increase that desire by making it possible to cover large areas with orthophotos without sacrificing framerate. So I would like to do some work on the texture pager, but I do not know if I'll be able to do that in time for 9.1.

My hope is to get 9.1 into beta early so that we can start getting third party aircraft authors involved with trying new features. If you are working on an advanced airplane, keep in touch!

Monday, March 10, 2008

Version 9 - Probably Too Late

If you have a third party add-on and you haven't tested it against X-Plane 9, well...honestly it's probably a little bit too late to fix compatibility bugs now. We're in RC, and only changes to fix critical bugs are going into the sim; everything else will have to wait for 9.1.

So if you have an add-on and you haven't tested it against 9.0, for crying out loud, do it now! Submit the bug anyway - if we can't fix it for 9, we can fix it for 9.1. But...after almost 14 weeks in beta, we've got to close this build up.

BTW a minor side note on fuel pumps...

In X-Plane 864, an airplane will run even if the electric fuel pump is off. (However, if the fuel pump failure is triggered and the electric fuel pump is off, either by switch or electrical-system failure, then you're out of fuel.)

X-Plane 900RC1 restores this behavior; in beta 25, an electrical failure would turn off all sources of fuel. This is wrong for a plane like the C172 where gravity can feed the engine.

In the long term X-Plane does need a more complex abstraction (e.g. does this plane have gravity feed or engine drawn fuel, or engine driven pumps, etc. etc.) but for now the 864 model, while inaccurate and incomplete, causes less problems than beta 25, which was simply wrong for a number of planes.

Friday, March 07, 2008

Installer Theory

Let me go into the installers in a little more detail before I start a mini-riot. Basically there are three "installation" operations that we (Laminar Research) support:
  1. Installing a new full sim from a DVD.
  2. Installing a new demo from the internet.
  3. Updating any installation from the internet.
The problem with the current installer model is that tasks two and three (get a new demo, update an existing installation) are handled by a single application. This means that the web-demo-installer/updater cannot know which choice (2) or (3) the user is trying to do (since both are legitimate) and therefore user error can result in the wrong operation.

The most common problem we have is users installing new demos when they meant to update a DVD install. The result is an old copy of X-Plane with scenery, a new copy without scenery, and a very confused user who has to contact us for help.

The solution I am working on is simple: provide separate applications for all three tasks.
  1. Like before, the DVD installer comes on your DVD. It installs the DVD, nothing else.
  2. A demo installer does nothing but install the demo. You get the installer from the web, probably from the "demo" page (which would have to no longer be an "update" page too). The demo installer always does a full install, and will stop you from trying to install on top of existing installations.
  3. The updater is always fetched by the app and updates the app. Thus the paradigm is that the app is "self-updating" (even though most of the work is done by a helper application). That updater (fetched by the sim) can only update the copy of the sim that fetched it.
Dealing With Problems

The above work-flow is meant to address most users doing the usual thing without technical problems. We must also consider how we will deal with tech support problems, and finally what the impact is on advanced users. There are two cases where we sometimes have to help users out:
  • If the DVD installer for some reason will not work for a user's computer, we usually provide a downloadable DVD installer. You would insert the DVD, run this "DVD installer" and thus install the DVD. This is not the normal case, but we will probably continue to provide DVD installers specifically for users with tech support problems (based on the tech support incident), just like we have been in the past.
  • We will provide the updater directly for users who need to update the sim but cannot launch the sim. This is the exception case, so a direct link to the updater would not be globally advertised on the main page of the sim. Rather it would be a support link, again only provided to users who really need it. 98% of the user base will be able to update from the sim.
When you run the updater from the web (the support case) it will prompt you to locate your X-Plane folder if needed.

Advanced Users

Will you be able to keep multiple copies of X-Plane around? Absolutely. But you'll have to manage your operations in terms of the "three operations" (make a new install from the DVD, make a new demo install from the web, update any one version you have laying around).

If you want to get a new web demo, you'll have to use a separate application than you would for an update. I don't think this is a huge problem; generally your best bet for keeping an old copy of X-Plane (when running a new beta) is to simply copy the entire X-Plane folder, rather than downloading the contents from the net.

Wednesday, March 05, 2008

Fun With Installers

So first, if I promised you scenery tools about now (MeshTool, AC3D, or any of the other things I've been asked for), please wait two weeks and ask again. I thought Austin would be cutting master DVDs while I would have time for tools, but instead I will cut the DVDs, so tools will have to wait until the masters are done.

To calm any fear: if you have X-Plane 9 DVDs, you will not need to buy new ones. We will be periodically updating our master DVDs (just like we always have), but we've taken a number of steps to ensure that everything new since the original "Beta 1" DVDs will be deliverable by reasonably small incremental web updates.

If you are reading this blog, you're probably an advanced X-Plane user, possibly one who makes airplanes or scenery or even plugins, and you're probably comfortable with computers and know a lot about the sim. Please bear in mind as I describe some of the changes we're making in the installer that we have users who have never used a computer before, and purchased their first computer to run X-Plane.

Simply put, the primary goal of the installer is to allow the non-experts to use X-Plane too.

Anyway, with this post I'd like to call attention to a feature that I believe very few people know about (because it was broken for a long time and yet we didn't get many reports).

You can update X-Plane 8 or 9 from the About Box - when you pick "About X-Plane" the sim will check our update servers for new versions, and if one is found, it will give you an update button. This update button will then download the very latest installer and launch it, setting it to update the copy of X-Plane you were running.

X-Plane 9 beta 25 will be out soon; when it comes out, please try updating by using the About Box in beta 24. Let me know if the process works correctly.

(What's so great about updating from the about box? Well, one of the biggest tech support calls we get is users who do not successfully patch their existing copy of x-plane, but use the web updater to fetch a whole new X-Plane, which of course has no scenery. When you get the installer from the about box, you always get the latest correct installer, and the sim tells the installer which copy to live. In the future we'll be removing the "destination" button when you run this way, so it's as simple as "go".)

Sunday, March 02, 2008

Don't use ATTR_cockpit outside the cockpit objects

I've blogged about this before, but...let me be totally clear:

Don't use ATTR_cockpit in objects that are not one of the two cockpit objects for your airplane.

Don't use ATTR_cockpit in the attached misc. objects for your plane - move the parts of the mesh that require ATTR_cockpit into the cockpit object.

Don't use ATTR_cockpit in scenery objects.

The OBJ spec basically says as much when it says "don't use cockpit features" outside of cockpits.

Now what goes wrong if you violate this varies with the betas vs. X-Plane 8, but I can tell you this: no version of X-Plane has ever shipped that will correctly handle ATTR_cockpit in attached objects for all cases.  There's always been bugs in this not-such-a-good-idea code path; it's just the severity has varied over time.

Hardware Profiles

X-Plane 9 currently recognizes (roughly) three categories of graphics hardware:
  • Non-Pixel-Shader hardware (GeForce 2,3,4, Radeon 7000-9200)
  • First-Generation Shader Hardware (GeForce FX 5nnnn, Radeon 9500-9800, X300-X600)
  • Later-Generation Shader Hardware (GeForce 6, 7, 8, Radeon X200, Radeon X700+)
That first bucket is pretty simple: those cards don't support programmable pixel shaders (as we know them today) and can't run any shader effects. The "use pixel shaders" check box doesn't appear in the rendering settings.

The distinction between the later two is a little bit more subtle. Basically the first generation of pixel shader cards (the 9700 and friends) support only 96 instructions for each pixel shader; this puts us right on the edge of sometimes not being able to draw all of our effects; we have to simplify the water slightly to make it work. The next generation of chips (X850 and friends) doesn't have this limitation.

By comparison, while NVidia cards have been able to handle long shaders from day one, the GeForce 5's shader performance is really poor.

So we bucket all of these chips as "first-gen". When we detect this we:
  • Simplify shaders slightly (gets us out of trouble with the 9700).
  • Don't default to shaders being on when the sim is first booted (because the framerate will probably be unusably slow).
Even though the 9700 provided very usable shader performance in its day, by the standards of modern GPUs, this older chip isn't that fast, so it's probably for the best that we not enable reflective water by default on machines with these cards.

By comparison, X-Plane deals with almost all other capabilities on an a-la-carte basis; particualr features are enabled if the right menu of hardware features is available. We do this to try to deal more flexibly with the wide variety of cards that are out there. Some examples:
  • You'll get hardware accelerated runway lights if your card supports pixel shaders and sprites (virtually all shader-enabled cards have sprites).
  • You'll get sun-glare effects if your card supports pixel counting (virtually all modern cards can do this).
  • The non-pixel-shader rendering code will show more detail if your card supports more texture units (this is only an issue with very old hardware).
I've been looking over hardware profiles a lot lately, and I suspect that the next big "jump" in hardware will be the DX10-compliant cards (GeForce 8, Radeon HD). There's a lot of fine print in what the various cards can do between all of the pre-DX10 cards; at some point when we decide what menu of features we'll require for rendering, we need to simplify.

My guess is that when we start to have "really advanced" pixel shaders that require hardware more sophisticated than what we need now, we'll simply require a DX10 card. Otherwise we'll have to sort through 8 different profiles of fine print, only to attempt to partially support cards that probably won't be fast enough anyway.

(That is to say, a feature is only useful for us if it can run reasonably quickly. It doesn't make sense for us to try to make a special "simplified" version of a rendering feature for, say, the X850 if every X850 is going to turn it off every time for framerate reasons.)

If any of this turns into hardware buying advice, I suppose it would be this:
  • If you are deciding between a DX10 and DX9 card (e.g. between the HD2400 and X1900, or GeForce 8600 vs 7900) go for the newer generation DX10 ards (HD or GeForce 8); if the card has decent performance you'll also be setting yourself up for future features.
  • As always, pay attention to the fine print of the model numbers, particularly the configuration. Lower model number cards basically have fewer parallel components than higher number ones and that leads directly to lower framerate.
I see an add online for the GeForce 8500 for $70 and 8600 for $90. But if you look at the links below, you'll see that the 8500 has only half the shaders of the 8600 - that's going to be a huge performance difference for $20.

(So the moral of this story is: try to get an HD or GeForce 8 card, but don't dip into the really low end cards because they're stripped down too far for X-Plane use.)

These pages (NV, ATI) on Wikipedia list specs for a whole pile of cards and can be useful to decode the fine print.

Sunday, February 17, 2008

Road Trip

I'm going to be more or less out of the office next week - the Amichai Margolis band is playing a series of shows in Florida. I'm going to have to take the laptop - we're too close to going final to miss a week of bug reports, but hopefully the next beta will stick for a while. (Initial reports indicate that the video driver initialization changes we made are fixing crashes and not causing new ones.) There will be a beta 23 shortly to address other bugs.

I wanted to take a moment to thank all of the users who have helped me debug video card compatibility problems remotely. I now have seven installed operating system configurations in my office, and it isn't nearly enough to see all of the problems we hit in field. Looking back at my in-box this week is a reminder of how patient you all have been in trying test builds, sending log after log, helping get to the bottom of some very tricky issues.

Things are going to be a bit busy for the next few weeks - once I get back we'll be finishing up the last few bugs, so it will take a little while to get back to questions about scenery, plugins, etc.

Thursday, February 14, 2008

Instability in Version 9

One of the reasons why the X-Plane 9 betas have had so many more crash bugs than version 8 is that we introduced loading DSFs on a second core. This feature makes scenery loads much slower and (by using the second core) impacts fps less while they happen.

The problem is that I'm still fumbling with code that will allow this in all cases. (That would be three operating systems, two hardware vendors, and a myriad of drivers, some new, some quite prehistoric.)

Beta 22 will be out soon, and will contain the fourth major rewrite of the OpenGL setup code for X-Plane 9. So far the initial tests look good, but we never know until we let a lot of users try the code and find the new edge cases.

It's relatively easy to tell if your instability is related to the use of OpenGL with threads: simply run the sim with the --no_threaded_ogl option. If things become a lot more stable, it's a threaded GL problem. Mind you --no_threaded_ogl is more of a diagnostic than a workaround; without threaded OpenGL, the sim will pause when loading scenery.

(Also, to clarify, you'll find talk on discussion groups and game forums about "threaded drivers". Threads are a programming abstraction that can utilize multiple cores. What I am talking about is X-Plane using multiple threads to load scenery - in our case this requires interfacing with OpenGL. But a threaded driver is different - it's just a graphics driver that's been optimized for multcore machines. These two concepts are totally different; you don't need a threaded driver to use X-Plane 9, and a threaded driver won't make X-Plane 8 load without pauses.)

Thursday, February 07, 2008

GeForce 7 and Water Performance

A number of Windows and Linux GeForce 7 users have discovered that the command-line option --no_fbos improves their pixel-shader framerate a lot. Windows and Linux Radeon HD users have also discovred that --no_fbos cleans up artifacts in the water. Here's what's going on, at least as far as I can tell. (Drivers are black boxes to us app developers, so all we can do is theorize based on available data and often be proved wrong.)

Warning: this is going to get a bit technical.

FBO stands for framebuffer object, and simply put, it's an OpenGL extension that lets X-Plane build dynamic textures (textures that change per frame) by drawing directly into the texture using the GPU. Without FBOs we have to draw to the main screen and copy the results into the dynamic texture. (You don't see the drawing because we never tell the card "show the user".)

We like FBOs for a few reasons:
  • Most importantly, FBOs allow us to draw big images in one pass even if the screen is small. For example, if we have a 1024x1024 dynamic texture but the screen is 1024x768, then withou FBOs we have to draw the image in two parts and stitch it together. That sucks. With FBOs we can just draw straight to the texture and not worry about our "workspace" being smaller than our texture. This is going to become a lot more important for future rendering features where we need really-frickin' big textures.
  • It's faster to draw to the texture than to copy to it.
  • If you're running the sim with FSAA, then we end up using FSAA to prepare all of those dynamic textures. In virtually all cases, we don't need the quality improvements of FSAA, so there's no point in taking the performance penalty. When we render right into the texture, FSAA is bypassed and we prep our dynamic textures a lot faster.
Since copying to a texture from the screen predates these new-fangled FBOs by several years, most drivers can copy from the screen to the texture very quickly; however we have hit at least one case where FBOs are much faster than copy-from-screen. That's really a rare bug, and as you'll see below, we see more weird behavior with FBOs.

When do we use FBOs vs. copying? Well, it's pretty random:
  • Pixel shader reflective water and fog use FBOs.
  • Cloud shadows and the sun reflection when pixel shaders are off do not use FBOs.
  • The airplane panel uses FBOs if the panel is 1024x1024 or smaller; if the panel is larger than 1024x1024 we draw from the screen and patch things together. So the P180 and the C172 are using different driver techniques!!
When you run X-Plane with --no_fbos, you instruct X-Plane to ignore the FBO capability of the driver, and we use copy-from-screen everywhere.

Mipmapping

There is one more element: mipmapping. A mip map is a series of smaller versions of a texture. Mipmapping allows the video card to rapidly find a texture that is about the size it needs. Here's an example: imagine you have a building with a 128x128 texture. If you park your plane by the building, the building might take up about 100x100 pixels on the screen; your 128x128 texture is a good fit.

Now taxi away from the building and watch it get smaller out your rear window. After a while the building is only taking up 8x8 pixels. What good is that 128x128 texture? Its' much too big for the job. With mipmapping, the card has a bunch of reduced-size versions of your texture laying around...64x64, 32x32,16x16, 8x8, 4x4, 2x2, 1x1. The video card realizes the building is tiny and grabs the 8x8 version.

Why not just use the 128x128 texture? Well, we'd only have two options with this texture:
  1. Examine all 16384 pixels of the texture to compute the 64 pixels on screen. That sucks...we're accessing VRAM sixty four times for each pixel. Accessing VRAM is slow, so this would kill performance.
  2. Simply pick 64 pixels out of the 16384 based on whatever is nearby. This is what the card will do if mipmapping is not used (because option 1 is too slow) and it looks bad. Thsoe 64 pixels may not be a good representation of the 16384 that make up your building side.
So mipmapping lets the video card grab a small number of pixels that still capture everything we need to know about the building at low res.

We don't mipmap our dynamic textures very often; the only ones that we do mipmap are the non-pixel-shader sun reflections and the pixel-shader sun reflections.

ATI

As far as we can tell, the current ATI Catalyst 8.1 drivers do not generate mipmaps correctly for an FBO-rendered texture. This is why without --no_fbos ATI users on Windows or Linux see very strange water artifacts. --no_fbos switches to the copy path, which works correctly.

At risk of further killing my track record of driver bugs in v9, we do think this is a bug. We have good contact with the ATI Linux driver guys so I have hopes of getting this fixed.

nVidia

It appears that the process of creating mipmaps for FBO textures is not accelerated by the hardware on the GeForce 7 GPU series. This is why GeForce 7 users are seeing such poor pixel shader performance, while GeForce 8 users aren't having problems.

Now poor performance is not a bug; there's nothing in the OpenGL spec that says "your graphics card has to do this function wickedly fast". Nonetheless, what we're seeing now is unusably slow. So more investigation is needed -- given that the no-FBO case runs so quickly, I suspect the hardware itself can do what we want and it's just a question of the driver enabling the functionality. But I don't know for sure.

Wednesday, February 06, 2008

The Limits of Orthophotos and Meshes in X-Plane

I get asked a lot about the limits of meshes and orthophotos in X-Plane. I'll try to answer this, but the answer isn't as simple as most people expect.

Texture Limits and Orthophotos

The maximum single texture size in X-Plane 8 is 1024x1024, and in X-Plane 9 it is 2048x2048.

I believe the maximum number of unique custom orthophotos that can be attached to a single DSF is at least 32768.

In practice, that number is pretty useless because X-Plane loads all textures for a DSF at the highest user-allowed res when the DSF is loaded. That means you tend to load a lot of textures. Every system is different and drivers have a lot to do with RAM efficiency, but generally you'll run out of virtual address space and crash the sim before you can attach 32768x2048x2048 of pixels.

X-Plane has no limits on how the texturing is applied - that is, you can use your 2028x2048 texture to cover an entire tile or a single meter. So again, the limiting factor on the resolution of your orthophotos is how much total area you want to cover and how much RAM you can spend (remember RAM is also used for mesh complexity, 3-d models, etc.).

You do not need to have enough VRAM to hold all loaded orthophotos; the video driver will paeg the textures into VRAM. Virtual address space is the limiting factor. How far you push it depends on a lot of subjective things:
  • If you expect your users to also run with a lot of trees, 3-d objects, cars on roads, and some plugins, you can't use a lot of RAM.
  • If you expect your users to have /3GB in their boot.ini and use nothing but your add-on, you can use a lot more RAM.
Generally the size of the DDS texture on disk is a good proxy for the virtual memory that is required to hold your textures.

It should be noted that these limits on texturing (due to X-Plane blindly loading a lot of stuff at once) affect all scenery types: objects, draped polygons, very complex airplanes, plugins, and not just terrain mesh orthophotos.

Getting Past the Texture Limit

It will take a future extension to the rendering engine to get past the current limits. Basically X-Plane will have to load textures at lower resolutions when they're farther away. I don't know when that is coming, but when it happens, it will increase the total amount of image data a DSF mesh can contain, because the limiting factor will be how much data is in the small area the user is looking at (since the rest can be stored at much lower res for far-away views). At that point the limiting bottleneck will be resolution (smaller means more data at once), not total image data.

Mesh Limits

Unfortunately, limits to the mesh are even more vague than limits to texture usage. X-Plane uses an adaptive mesh - basically you can put your vertices wherever you want. So the highest resolution you can achieve might be much smaller than 1 meter resolution, but you can only do this for a small area before the total mesh size gets too big. But this is okay - the intention of DSF is to let you put a lot of detail where you need it.

I believe that once again memory provides the first limitation to the mesh. That is - you'll run out of memory loading your insanely huge mesh long before you hit a limit to the DSF container structure. And once again, even the RAM limit isn't a hard limit because that virtual address space is shared with texures. Your mesh density limits actually go down when your textures go up because it's a zero-sum game.

Estimating Memory

Here are some ideas on how to estimate your memory footprint:
  • Run X-Plane over ocean to get an idea of the baseline memory use that the sim needs without extra scenery.
  • Load your mesh without textures (move the textures away) to find the cost of the mesh itself. (I am going on the assumption here that you can rescale your mesh using whatever mesh generation tool you're using).
  • The size of DDS textures is a good proxy for the memory used.

Tuesday, February 05, 2008

MeshTool: The Seeds Are Sown

Last night I created the "seed" files for MeshTool. Let me explain what these files are and why we'll need them.

MeshTool is a wrapper around our irregular-mesh generation code. It will allow third parties to create base mesh scenery without having to create triangulations. Just like DSFTool saves people the work of having to encode DSFs (with point pools, command lists, and all that ugly stuff), MeshTool saves people the work of having to create their own triangulations.

MeshTool is a low level tool - you provide a text input file and some data. It's designed to be an engine underneath tools like PhotoSceneryX, not an end in itself.

MeshTool will create "default" land-use terrain that approximately matches the global scenery, water, and custom orthophoto-based terrain. You (or a program you use) provide a text file that describes the boundaries between custom photos, land, water, and airports. You must also provide a SRTM-style HGT file for elevation.

How does X-Plane know what land-use should go on what terrain? That's where the seed files come in. Our global scenery is generated from a set of rules that take into account morphology (land height and slope), approximate climate, and general land use. You provide the terrain shape via the HGT file, and we provide you with a seed file that contains climate and land use for that DSF tile.

Why do we provide the seed file rather than letting you find and create climate data? Well, our rules are tuned for a very specific pair of data sets; by providing the exact climate and land use data that we use, we assure that the rules files work correctly. The purpose of MeshTool is not to customize land use terrain, and we do not provide a mechanism for it. The purpsoe of MeshTool is to let you put orthophotos and new coastlines into the base mesh.

The good news is: seed files are tiny. They are typically 4 kb-8kb each; the entire data set is 322 MB total. That's because the climate data is only 3x3 per DSF and the land use is only 121x121.

I hope to get MeshTool into some kind of testing within the next few weeks; if you are a programmer and would like to feed MeshTool from your own program, please contact me and perhaps I can arrange an alpha copy. I will also post the seeds as soon as I can.

Monday, February 04, 2008

No Answer Does Not Mean Go Ahead

Every now and then I see a comment in an X-Plane forum somewhere to the extent of:

"Joe Author made this great scenery pack for FS2K4, I tried to contact him about a port. I got no response, and the pack is free anyway, so I've posted my conversion."

Simply put, you can't do that, at least not in the United States. Copyright law is very clear on this subject: if you don't hear back from the author, the default is that you do not have permission to create a derived work.

(The fact that the original package was "free", meaning cost zero dollars, is not at all relevant. The author retains his rights to his own work even if he doesn't charge money.)

A simple thought experiment reveals why it has to be this way: if I was giving away my new program as a promotional period and went on vacation, and you decided to post a derived work because (1) it was free and (2) you hadn't heard from me, I would have no way to stop an illegal use of my work that I did not ever want (nor ever indicated that I wanted). "Free" plus "no one is home" is simply not a high enough bar to protect authors.

Thursday, January 31, 2008

Algo-Gen

I may be fighting a pointless and unwinnable linguistic battle, but I have to try. People very often refer to the default city buildings in X-Plane as "auto-gen" but by any reasonable definition of "auto-gen" they are not really auto-gen at all.

Now these are all made up computer terms, so we can't really check the dictionary. But "autogen" scenery (short for automatically generated) usually refers to scenery that is created by a flight simulator itself, usually while you fly, and usually by placing 3-d detail in places that match the base terrain. This exists in FSX, and existed in X-Plane up to version 7.63.

X-Plane 8 doesn't have autogen!!!!!!! X-Plane 8 has scenery that is generated by computer programs, but X-Plane is not the computer program that is doing it. When you see a ton of buildings piled up in New York City, that is not becaues X-Plane looked at the New York city base terrain and said "hrm - some buildings would be nice."

What actually happens is we analyze New York City when we create the global scenery (before we ever burn the DVD masters) and the DSF generator places all of those buildings in New York City. X-Plane simply gets a huge list of buildings from the DSF and draws them.

I am going to try to coin the term "algogen" (algorithmically generated) to describe these buildings that (like autogen) come from a computer generating semi-random buildings from input data, but unlike autogen, algogen is a process that runs once before the scenery is made.

So how is algogen and autogen different?
  • You can't change the pattern of algogen building placement by editing files in the sim. The algorithm has already been run! You can replace the buildings using an overlay (that excludes the base) or by using a library of models to substitute models.
  • We are trading data size for computation. The DSF is bigger because it lists the location of every building in New York, even if they were just algogen buildings, but the job of placing those buildings is less difficult because X-Plane does not have to check each building against each road. That has been done in advance.
  • Changing the scenery via an overlay doesn't change the algogen! Add an airport via an add-on and you have to exclude the buildings. (But if you send that airport to Robin, the next global render will include it and the algogen will skip the airport automatically.)
Note one of the interesting results of algo-gen: X-Plane can't tell the difference between an alg-gen building and a hand-placed one! They're all just objects in a DSF. The fact that algo-gen buildings disappear with lower settings is because the sim/require_object property in the DSF header tells the sim which objects are important, and our generator always signals the buildings based on obstacle data as important. But algogen as a process is not visible to X-Plane!

And that's why I'm spending so many words on trying to distinguish between "algogen" and "autogen" - because the processes are fundamentally different, they're very different for scenery authors to work with. As a result, authors coming from X-Plane 7 or FSX will be very surprised if they try to understand X-Plane in terms of autogen....they won't be able to find the autogen config files, and the autogen buildings won't react to other scenery changes, because they're not actually autogen at all!

Algogen is a classic pattern of "precompute" vs. "compute-while-fly". Generally precomputing gives authors more flexibility (in our case, we have an obj engine that can handle a lot of objects, so authors can make their own objects of the same density as algo-gen with the objects placed anywhere) at the expense of making it more complex to edit the existing scenery (edit the mesh and the algogen doesn't change).

When we started the v8 scenery, two things pushed me toward precomputation:
  • In the past, changes in X-Plane's rendering engine had broken third party add-ons. So a precomputation strategy (by getting the scenery code out of the sim) means that the sim is doing less "interpretation" and thus the interpretation of scenery is less likely to change.
  • We wanted to focus on performance, which means getting computation out of the sim whenever we could.
Now that last point isn't quite as important as it used to be...when we were doing the design (during mid X-Plane 7), dual core for everyone wasn't on the radar, so the penalty for complex computation while flying is lower (and thus we have more expensive in-flight computation, like forests and completely draped bezier curve-based polygonal pavement).

But I think precomputation is still useful. Even with dual core, the algorithm that places X-Plane's algo-gen bulidings can take one to two minutes for a 1x1 DSF tile on a very fast computer. That's still a load time that's out of the question for us; even on the second core, the DSF wouldn't be "ready" in time for you to fly it. So one use of precomputation is to run algorithms that are more expensive than you can have in real-time. (That algorithm to pack objects inside an irregularly shaped polygon made by roads and land features is not fast.)

More importantly, precomputing does give us a nice advantage in the use of storage data. We ship about 50-60 GB of final scenery, but the source data is well over 100 GB. When we run the algogen algorithm, we have access to the full set of source data: coastlines, elevation, and land use before any simplification is done and any data is thrown out. So we have the potential not only to do a more complex analysis, but to do the analysis on a larger data set.

The down-side of precomputing is that if integration of all data is saved until sim time, there is the potential for third parties to contribute separate data to the sim via add-ons and still have the integration of those data sets work well. This doesn't always work out - see complaints in online magazine reviews about combining orthophotos and new road grids in FS2K4...they don't integrate because neither of those types of resources can be integrated to match the other in real time. But autogen still does a much better job than algogen at this; algogen basically has to be recut when other data changes. (And that is our intention - if you change the road grid, exclude and replace the objects too!)

Saturday, January 26, 2008

Performance Wrap-up (for now)

The story on X-Plane performance is never over, but the chapter that is 9.00 pretty much is. I think we'll be RC in the next build (if all goes well). Certainly a lot of the things that are still performance "problems" will require changes larger than we can do in a late beta.

I say problems in quotes because a lot of what's been reported lately is in the form of: a huge screen res + a lot of shaders + a lot of FSAA = slow fps. That's not really a bug, that's an engine limitation. Now I want to make the engine as fast as possible, and a lot of this pixel shader stuff is new to 9.0, so if our track record for tuning stays the way it was for v8, we'll probably get some efficiency improvements later.

But unfortunately there's an underlying limitation: the new water and fog both cause the rendering engine to consume significantly more hardware resources than it would otherwise. Turn them on and you get prettier pictures at a price.

Just to post a a few general things I've found:
  • X-Plane 9 will tell you where your GPU really stands. GPUs that were very adequate for X-Plane 8 (like the GeForce 6600 GT) will turn out to have nothing left in reserve for v9, while GPUs that were bored in v8 (the GeForce 8800 GTX for example) will show what it really has.
  • Generally the cost of going from no shaders to shaders with water reflections of "none" and no volumetric fog should be very low if your screen res and FSAA don't add up to something crazy (like 16x FSAA at 2048x2048).
  • If you do have serious performance hits, try --no_fbos in the command-line; some drivers seem to have trouble with them.
  • The P180's virtual cockpit is a lot more expensive than the other ones, because it has a huge panel that is used in 3-d. We'll hopefully rebuild the cockpit at some point.
  • Turning water reflections to "complete" is very expensive. Watch the water and use the lowest setting that looks good. You don't need complete reflections if there are a lot of waves!
  • Shaders, FSAA, and screen size are all pulling from the same set of resources - be careful about cranking up all three.
  • Check your v-sync - a lot of users whose vsync clamped them at 60 fps in v8 will be clamped at 20 in v9.
  • Do your testing with texture res set low, then crank texture res later; pixel shaders also require the allocation of VRAM that can't be purged (for things like reflection images) so running out of VRAM can show up in some weird ways performance-wise.
  • The new Intel iMacs have serious performance problems with shaders on. This is due to driver limitations; given the much better performance under BootCamp, I expect the Mac performance to get better when the drivers are updated. For now I'd keep shaders off.
For now, please hold off on sending me performance reports. I just don't have time to address them. In the future I will try to solicit very specific performance data points that we need to check. Perhaps in the future we can also set up a database of fps-test results to have a more comprehensive idea of how the hardware does its job.

I expect future features to appear in v9 that further eat hardware; those features will have an off switch. You may have to pick and choose what graphics you enable; there is no free lunch here. I also expect new graphics cards to emerge that make the GeForce 8800 GTX seem quaint!

ATI: 2. Ben: 0.

What a difference new drivers make. ATI's latest OpenGL drivers (Catalyst 8.1) seem to work quite well with X-Plane. On two fronts:
  • Linux. Turns out all you need to do to make X-Plane happy on Linux with ATI hardware is update the drivers. I'm running with the Cat 8.1 drivers on my MacBook Pro and things look good. Use Catalyst 7.11 drivers or newer! No more MALLOC_CHECK_=1 or --no_threaded_ogl. With the next beta, you won't have to use --force_run anymre.
  • Windows. We were getting reports of corrupt screens on startup, and with the Catalyst 8.1 drivers these reports became very frequent. Turns out our threaded OpenGL code was doing something naughty*. Beta 19 fixes this.
The only known issue I can think of is: if you see corrupt water reflections, run with --no_fbos.

* Well, the way you set up threaded OpenGL on Windows and Linux is not very well documented, so I say naughty in that we made the drivers unhappy. I have yet to find a document that states clearly whether what we were doing is correct or not. We had to guess.

Friday, January 25, 2008

2 Blog or Not 2 Blog

I figure there are three things that making blogging suited for software:
  1. A small interested group of users can subscribe to a blog without giving out an email address.
  2. The information in the blog can be found using a general search tool like Google.
  3. It's easy for a busy programmer to post.
I think this last point is not to be ignored - I post a lot here because it's easy enough to whip up a blog post that I can write one while the sim is booting.

The down-side of this is that blogs are not self-organizing. The blog is chronological, and somewhere within a heap of 200+ posts are detailed information on scenery topics not documented on scenery.x-plane.com.

That's not good. So I'll be trying to make a concerted effort to write real permanent documentation for some of the new scenery system topics that I cover. Documentation on DDS is in the works.

Part of the problem is that my interface for updating the X-Plane scenery website isn't that robust. One of the nice things about the plugin system being a Wiki is that it's easy to organize and easy to edit. (And one of my frustrations with "support forums" is that they don't self-organize...they mix questions and answers based on history and not a search key that a user might use, like "what's wrong with my card." We'll be supplementing the Linux forum with a Wiki soon.)

Airplanes - How it Fits Together

Here's a summary of the new airplane features in 9.0 (and some coming). Hopefully this will give you an idea of what new capabilities are available for modeling planes in X-Plane 9. This list will sound like a broken record - virtually all of these features are optional; you don't have to recut your finished airplanes to use them in version 9.

2-d vs. 3-d Panel

You may have noticed the new "3-d panel" option in PlaneMaker 9. This allows you to build a separate panel for the purpose of providing the texture to ATTR_cockpit (or ATTR_cockpit_region). You can then:
  • Provide alternate instrument artwork in a cockpit_3d folder. (This lets you have perspective artwork for the 2-d cockpit and orthogonal artwork for the 3-d cockpit.)
  • Pack your instruments together tightly to save space. (There is a real cost to large panels, so using a 1024x1024 panel for the cockpit object is a lot better.)
The 3-d panel is strictly optional, fully replaces the 2-d panel only for cockpit objects, and is activated by providing a custom panel background in a cockpit_3d folder. (See the "Example Plane-Widescreen+objects" plane in beta 19.)

ATTR_cockpit_region

Cockpit regions are an alternative to using the entire 2-d panel to texture your objects. They provide a few advantages:
  • Performance. By requiring a power of 2 and allowing you to use a sub-area of the panel, cockpit regions avoid a lot of wasted computing that ATTR_cockpit can cause.
  • Next-gen lighting. Unlike ATTR_cockpit, real 3-d lighting is applied to the panel when you use this attribute. This means that you will get a gradual decrease in light on your geometry (correct based on the angle of the sun) that matches the rest of the object.
Please note that you can mix and match which way you get your cockpit texture and whether you use the 2-d or 3-d panel feature (above) independently. However, you can only use ATTR_cockit or ATTR_cockpit_region in your airplane, not boht. ATTR_cockpit is still supported.

Generic Instruments

Generic instruments let you build instruments that follow some basic shapes (needles, tapes, etc.) that can be tied to any dataref. This both lets you customize particular instruments very precisely or create an instrument driven by a plugin dataref. These instruments are optional in version 9 - the old "premade" instruments are still supported.

New Datarefs

X-Plane 9 provides new datarefs targeted at airplane authors. The datarefs are better organized and have clearer names. But the old datarefs still exist, so legacy planes do not have to be updated.

Generally the entire cockpit should use only sim/cockpit2/ datarefs, and the plane exterior should use only sim/flightmodel2/ datarefs.

One special feature of these two sections: if your plane is used as an AI plane, these datarefs will animate the plane with the AI plane's control deflections, not the user's control deflections. So using these datarefs fixes the "AI animation" problem.

Plugins in Aircraft Folder

Version 9 airplanes may have a plugins folder (inside the ACF package) with fat plugins inside them. If you develop a plugin for your airplane, consider packaging it this way -- this will allow your users to install the airplane with a single unzip for all platforms and no extra "drag-this-file-here".

Plugins in the airplane folder is optional - you don't have to provide a plugin, and plugins that are installed in the main Resources/plugins folder will still work. Still, I encourage you to use this feature because it makes the install process a lot simpler. The X-Plane SDK website will have documentation on fat plugins.

Liveries Folder

X-Plane 9 features a new "liveries" folder. Liveries (replacement exterior paint for airplanes and their attached objects) can be placed in packages in the liveries folder to greatly simplify the process of repainting an aircraft. See the "Example Plane-Widescreen+Objects" for an example.

While the liveries feature is optional, I strongly encourage anyone doing repaints to adopt it. Liveries can be switched by the user in the sim without any file manipulation; there is thus no risk of accidentally deleting or breaking an aircraft.

Large 2-d Panels

In X-Plane 9, a panel can be up to 2048x2048 in size. You pick the dimensions. The panel will scroll horizontally if necessary.

Note that if you use the new 3-d panel feature, the 2-d and 3-d panel do not have to be the same size. I would recommend a large 2-d panel (to fill large monitors) and a smaller 1024x1024 3-d panel (for performance).

Hiding Parts

X-Plane 9 will allow you to hide aircraft parts. Many v8 planes use OBJs to model the plane geometry, and use a transparent ACF texture to hide the ACF. Setting the parts to "not drawn" saves the CPU time that X-Plane would spend drawing the airplane, and is thus more efficient.

Keyframes

X-Plane 9 supports key-framed animation; this is useful for the scenery system, but for airplanes it allows for much more complex and realistic animation. OBJs that don't have key frames still work.

Manipulators

This is a feature coming in the future: the ability to control how the user clicks and interacts with the cockpit object in detail. In X-Plane 9.0 we only support clicking on cockpit-textured geometry; manipulators will make features like draggable handles a lot more workable.

Global Illumination

X-Plane 9 does not yet offer a lot of control of the in-cockpit lighting environment; we'll be working on this in future versions. These features will be opt-in...that is, you'll have to change your model to get the new features, and old planes will work the way they always used to. It is likely that you'll have to use "modern" airplane-building techniques to use these new features (meaning OBJs, named or custom lights, lego brick instruments ,etc.).

Tuesday, January 22, 2008

Off Topic: Kid in a Candy Store

I try not to divert this blog from X-Plane related stuff, but I just can't resist. Go to the itunes music store and search for:

amichai margolis

Ami has been a friend of mine for years; Orah V'Simchah is the first CD he's recorded of his own music (I played bass on it). A few time a year he pulls a band together and we play a few gigs. Anyway, enjoy the tracks!

Friday, January 18, 2008

Reflections in the Wrong Place

We get a lot of bug reports showing strange reflections in the water. Some of these we can fix easily, and some will be more difficult, if not impossible. There are two fundamental constraints on the water-reflection code:
  • A reflective surface (read: water) must be approximately flat to be correct. This is just how the algorithm works. (If you want to see why a non-flat surface fails, try to draw a camera position opposite the reflection plane and trace rays through this "reflection camera". Draw reflected objects that are close and far from the water and then observe the parallax error you get if the reflection plane curves.)
  • We need one "water camera" for each flat reflection plane. You can't just statically offset when we have multiple elevations. (When drawing your diagrams, note how an elevation change causes a change in reflected angle, not just an offset.)
So the water will always have two problems: the earth is round (so nothing is really flat), and we can have lakes of multiple elevations (and we can't afford to render a water reflection per lake).

X-Plane tries to get around the non-flat water problem by picking little bits of the water that are flat (and seem useful) and using them to define reflections. This algorithm will always have problems, but at least it can be tuned.

Now there are also some things that we can fix with the water:
  1. The math in beta 18 is simply wrong, something that will be fixed in beta 19.
  2. The ocean is built from polygons that are too large; this introduces approximation errors when we try to pick "a little bit" of water to use to figure out our reflection plane.
There is one more problem that I see, especially in airports like PAKT: if there are two water surfaces of different heights that are nearby, X-Plane uses a slanted water plane that tries to include both. This works very badly - the resulting slanted plane doesn't look even remotely plausible. I'm not sure how soon we can tune this problem.

Why You Can't Have a Setting

I get a lot of requests for settings...the email is typically something like:
  • Some feature in X-Plane is defaulted to X.
  • I like it better when it's like Y.
  • Can I have a setting to change the feature between X and Y.
Raymond Chen has a great posting that I think is very topical: "In order to demonstrate our superior intellect, we will now ask you a question you cannot answer."

This brings up one of the main reasons why we shy away from more settings: the more complex we make X-Plane's configuration, the less likely it is that the average user will be able to set the sim up correctly. Settings requests usually come from our most advanced users, but we also have users who have never used a computer before. Really! I've been on the tech support calls - they are very nice, but way overextended on the computer side of things. Should we allow them to pick whether scenery geometry is store in AGP memory vs. VRAM?

From our perspective, having a setting that a user doesn't understand is worse than neutral, it's actually harmful. Every one of those settings is something that can go wrong with the sim. I removed the ability to set the level of detail bias to positive (in other words, extend the visibility distance of scenery beyond its original design) after about 500 complaints of "low framerate" from users who had maxed this setting out (causing a 4x increase in 3-d processing load) without knowing (1) what the setting was, (2) what it was good for or (3) what the down-side was.

Could we present all the info to make intelligent decisions on the rendering pages? Honestly, probably not beyond a certain point...we would devolve our sim into a lecture on working sets, bottlenecks, and the OpenGL pipeline long before the user got flying. (Wait, that's my blog! Doh!!!) At some point the sim just has to do its best to do the right thing, or something similar to it. Just as Raymond points out that the default answer to any dialog box is "cancel", the default answer to any rendering setting is "all the way up."

When I tell a user who wants a setting that he or she can't have a setting because some other user will abuse it, the answer is almost always: well why don't you have two settings screens, a simple and advanced mode?

Besides the irony (of trying to solve the problem of too many settings with another settings), Raymond also points out that no location to hide an advanced setting is ever quite good enough. This is something we have struggled with, choosing command-line options more for to pragmatic reasons than because it's a great solution.

This doesn't mean you can't ask for command-line options...I am just trying to point out some of the thinking on the other side of the coin.

Wednesday, January 16, 2008

When Can You Not Use DDS?

There are a few cases where you cannot use DDS files in X-Plane:
  1. Airplane 2-d panels (any layer - base, lit, -1 shadow layer, 2-d or 3-d).
  2. Airplane instrument images.
  3. Bitmap-based region specification referenced in a library.txt file.
  4. Any gray-scale/alpha-only texture (e.g. mask files in the scenery system).
Beta 17 is treating cases 1 & 2 as an error; beta 18 will simply stop looking for DDS files in those cases.

Please note that airplane panels and instruments are not compressed right now, so there would be no performance benefit to using DDS in these cases. (If anything, PNG has smaller file size when compression is not used.) If we ever allow compressed panel textures, we'll probably allow DDS panels at the same time.

Case 3 is just a particular version of case 4 - that is, the region bitmap is black and white (1 channel) so DDS provides no benefit. Use a gray-scale no-alpha PNG!

Tuesday, January 15, 2008

Simple Optimizations for Airplanes and Objs

Just a few basic things:

For airplanes where you don't want to show the PlaneMaker part (because you've rebuilt the plane visually using OBJs):
  • In X-Plane 9, set the parts to be invisible - this is faster than drawing them with a transparent texture.
  • In X-Plane 8, if the texture is transparent, please downsize it! A 1024x1024 texture that's fully transparent is just a waste of VRAM!
For any version, any object, avoid using ATTR_diffuse_rgb when possible. You can get the same effect by tinting your texture and save unnecessary state thrash.

Saturday, January 12, 2008

Panels Will Be Opt-In

First, the most salient point: in X-Plane 9.00 beta 16, 2-d panels and 3-d cockpits should both look the way they did in version 8. That is, your v8 plane should look good in v9 without modification. This is due to both:
  • A bunch of bug fixes regarding burn-in and night lit layers.
  • "3-d" lighting is not applied to the cockpit texture.
On this last point, my hope had originally been that I could apply 3-d lighting to the cockpit and simply make existing content look better. It has become painfully clear to me that this is not possible -- existing planes are very carefully tuned to look good despite what I can only describe as "inconsistent" lighting rules in v8. Applying more consistent general 3-d lighting wrecks this tuning.

So new features regarding 3-d cockpits will be opt-in - that is, you will have to change your model to start using them. Existing content will work the way it used to.

I will explain what new features are available in 9.0 and what will come in 9.1 in a separate post real soon.

Water: What's Coming Now

Beta 16 should be out relatively soon and includes what are probably the last of the water improvements for X-Plane 9.0. (All future water improvements will be in free patches or future releases.) Here's what made it into this particular version:
  • Water wave patterns now follow the wave settings in the weather-water screen.
  • "Fetch" data from the DSFs (information about whether water is subject to ocean waves or not) is used to modulate wave patterns - go to KSAN and set the wave height fairly high. Notice how the waves die out to ripples in some of the bays.
When I last blogged about water, I hadn't realized that we already have "fetch" data in the DSFs (and have since 820). Bear in mind that the fetch data in the current DSFs is not very good - I am sure you'll find plenty of strange cases where there are ocean waves inland, or calm areas in the ocean. The algorithm just isn't that good. New fetch data only comes with new global scenery.

The physics engine is not synchronized with the visual waves, although they both use the same wave height from the weather screen. The physics engine isn't using fetch yet. I may be ale to address this in a future beta, but I'm not sure.

What I have had to punt on (for now - remember, a ton of new features come to X-Plane in the form of free patches!) is the issue of reflectivity, sparkle, and water color from various view angles. As I explained in my previous post, the way the wave pattern is filtered with view-distance causes the water to look unrealistically calm from far away. Now that we can haev taller waves this is less of an issue (set very high waves to avoid the reflective look as much as possible) but it's still there.

I have some possible technology ideas on how to address this problem, but we don't have enough time left in this build to go all the way with them.

One final note on water: Windows users witih Radeon HD hardware can work around the corrupt reflective water issue by running the sim with --no_fbos.

Wednesday, January 09, 2008

X-Plane Water - Now and the Future

Randy forwarded me a very detailed email message from the features list about water. First a few notes:
  1. I don't read the features list - I'm only blogging this because Randy forwarded it my way.
  2. Procedural water (that is, procedural waves with the sky as a reflection) is a default shader option because when we first looked at it, it seemed to be "low cost". Some hardware really chokes on this option. But one trend that's clear: I have not seen any hardware that can do volumetric fog but has trouble with water. When it comes to expensive computations, vol-fog is the the heavy effect. If your card can run with vol-fog even remotely well, you're not going to get any kind of fps boost by turning off water. So the question of whether procedural water can be optional is one of whether the water looks good (which I will discuss below), not one of performance.
  3. I've received plenty of emails about how we can "cheat" on the reflections to make them faster. To be honest, these ideas aren't very useful...you really need to know how the rendering engine works on a low level to find good ways to cheat on reflection quality.
(It's not enough to make the reflection texture less good looking, you have to do so in a way that makes it take less time to render! We've basically already taken all the optimizations we can - that's what that water reflection detail popup does. Keep it on a lower setting.)

So with that in mind I'd like to bring up three issues with the reflective water and give you some idea of the roadmap. I should say that when I list features as coming in the 9.xx or 10.xx time frame what I really mean is: some depend on new global scenery and some do not. It's possible that we may do a ton of water work in 9.1 or we may not do any more for five years. I can't really make a good prediction on future features after 9.0, except that they'll be, um, really cool. :-)

Water Weather Settings (9.00)

X-Plane currently provides a global wave height setting. X-Plane 9 beta 15 ignores this setting and simply creates calm water. This is simply a link-up between the physics and the shader that I had not gotten to until now. I believe that beta 16 will address this; the water wave height will match what you dial in. This still is only a baby step, but there will at least be a workaround for the mos common complaint (the ocean looks like glass) in that you can set the wave height to 10 meters.

Water Properties (9.xx, 10.xx)

Now Peter brings up an important point: the properties of how the water look vary with their location. First I must point out an architectural issue: you can't do a great job of computing "fetch" (open runs where wind builds up waves) in the sim because the area adjacent to the current water may not be loaded. That is, if X-plane doesn't have the pacific ocean loaded, how can it know that the waves hammering San Francisco can be pretty big? So I have always viewed fetch as something to pre-compute into a DSF mesh.

(This does bring up an architectural issue...how can we have properties on open ocean with no DSF mesh? The answer might end up being that we do have to provide water-only DSFs for bays and inlets that are large enough to cover a whole tile.)

Now it turns out that (secretly) the DSFs have contained a very crude version of fetch since 8.20. In version 8 we didn't really have the shading power to visualize it (I did have some experimental code once) and in version 9 it is not yet hooked up. The fetch calculation isn't very good, but it does exist. In the long term, we've talked within the company about including bathymetric data (water depth) and even water properties like clarity and the color of the goo in the water. Improving the water metadata would be a next-global-scenery feature, and we don't recut global scenery very often.

(The DSF format is flexible...you can encode just about as much extra data into the mesh for water as you want - it's not a format change.)

So I think you may see some improvement in the water as we utilize existing fetch data in the mesh, and some improvement as we encode more meta-data into the mesh. Note that to do any of this we need to change the sim a bit...right now it assumes a constant wave height - this would no longer be true. I think these kinds of improvements will start during the 9.x run but not be in 9.0.

Filtering Errors (9.xx)

This is the most technical issue with the water, relating to how the graphics hardware works. The problem is that the way the far-view of the water is computed is too reflective due to down-sampling.

In real life, if I am in front of a body of water with 6 inch waves, I will see two things:
  1. Near me, I will see the waves themselves. Part of the waves will be dark, because their surface normal faces me, so the Fresnel equation says I see down into the water, where I see the bottom (or if it's deep enough, darkness). Part of the waves will be at an angle to me and act reflective, picking up colors from the sky and maybe surrounding terrrain. So my waves are going to be a mix of the sky color and some kind of dark color, with the color contrast allowing me to see the shape of the wave.
  2. Farther out, the waves will be smaller than my eyes can distinguish and I'll start to see a more consistent water color, which is a mix of all of the various sky color inputs and darkness. The particular mix might depend on the chop and shape of the waves and angle to the sky. The important thing is: there is scattering of the reflection at a level smaller than my visual acuity, so I see sky color, but I don't see a sky reflection, and that color is darkened by the Fresnel equation.
Now in X-Plane the problem is this: the water waves are built up procedurally from a noise texture. As we get farther away, that wave texture is reduced in quality. Unfortunately, the graphics hardware averages together the wave shape, not the resulting color from the wave shape.

So instead of getting sky + deep = darker blue for the water, I get peek + trough = flat water! In other words, at a far distance the waves are canceling each other out before color lookup, giving us a perfect mirror in the far-ground.

This is fundamentally an implementation problem - I bring it up only because it's a counter-intuitive one. In the immediate future, the "glass lake" problem will become less because filtering only kicks in like this when the waves become less than one pixel - with the option for taller waves coming, the waves should be visible farther out. In the longer term we'll probably put in new shading code to address filtering problems.

Reflection Positioning Bugs (9.0, 9.xx)

As of beta 15 I thought I had fixed most of the reflection positioning bugs (that's what happens when something reflects in the wrong place in the water) - the geometry for this is made complicated by the Earth being round. I don't expect to nip all of these bugs in 9.0 but I do hope to get most of them, and I will keep working on this as bug reports come in.

Wave Shape (9.xx)

Finally, our wave shapes are quite primitive - it's just shaped procedural noise designed to look tolerably like water. We have a framework into which we can insert more complex wave equations (at the cost of some framerate). I don't know what the future will bring in this area. The v9 water sets out a new foundation onto which we can do more complex water. But we have to crawl before we can walk.

Tuesday, January 08, 2008

A New Broken Record

For years now I've been harping about ways to keep the number of batches down in your scenery. A batch is a single submission of triangles to the graphics card for drawing. Batches get rendered fast even if they contain a ton of triangles, but changing modes between batches is not very fast, so a few large batches is hugely better for performance than a large number of tiny batches.

To play that broken record one more time, there are two ways that you (a scenery designer) can cut down the number of batches):
  • Use a small number of larger textures instead of a large number of small textures, preferably sharing textures between similar scenery elements that are placed nearby. X-Plane will do its best to merge the content that uses those textures into single batches. We call this the "crayon rule."
  • Use less attributes in your objects. Attributes usually require a new batch (after the graphics card mode has been changed due to the attribute). So if you've got 1000 attributes in your object, you've got a problem.
Well, with X-Plane 9 I have a new broken record: avoid overdraw!

Overdraw is the process of drawing pixels on top of other pixels on screen. It happens any time we use blending to do translucency, and any time we use polygon offset to build the image in layers.

Overdraw is bad because with X-Plane 9's pixel shaders, most users are slowed down by the graphics card's ability to fill in pixels (pixel fill rate), with those complex shaders being run for every pixel. If you are at a screen-res of 1200x1024 looking at the ground with no objects, that might be 1.2 million pixels to fill. But if there is an overlay polygon covering the ground, we have 2.4 million pixels to fill! That's a huge framerate hit.

Right now there's not much you can do about overdraw. Once MeshTool comes out I will post some guidelines on how you can limit overdraw.

We took a step in the v9 global scenery to limit overdraw: in X-Plane 8 the global scenery tried to hide repetition of flat textures by drawing them over each other with offsets. In X-Plane 9 this is done in a pixel shader (e.g. the texture is analyzed and swizzled in the shader and then drann once), cutting down the number of times we must draw.

If you turn pixel shaders on and off in a flat area like Kansas you might see this if you compare the screenshots - the farm textures are more repetitive without shaders. This gives faster fps to everyone (with or without shaders) by eliminating overdraw.

Monday, January 07, 2008

Scenery and the New Plugin SDK

Sandy and I are working on a major revision to the plugin SDK (all the old stuff will work, we're just adding new things) that should be available in X-Plane 9 soon. The new "2.0" SDK APIs include some new functionality for working on scenery.
  • Plugins can find the height of the ground at a given location, which is necessary to draw in the 3-d world in a realistic way (e.g. vehicles that drive on the ground in a sloped airport environment).
  • Plugins can load and draw OBJ files using X-Plane's built-in facilities. I've posted OBJ drawing code in the past, but this makes things even easier.
  • Plugins can lookup virtual paths in the library to find artwork from scenery packages.
This makes a number of scenery-system concepts available to plugins.

I've been resisting OBJ-drawing support in the SDK for a while, but a few things changed my mind:
  • We've moved as much drawing in X-Plane to OBJs as possible and it's been a big win. A lot of the dynamic elements are OBJs, they're used in scenery and cockpits and airplanes. Using OBJ files means our artists (who are not programmers) can customize just about every aspect of the sim. So by providing a file format with a rich tool chain to plugins, hopefully we are helping third parties streamline content development.
  • With pixel shaders, X-Plane's 3-d drawing environment has become complex and hard for third parties to safely augment. By encoding drawing at a higher level via pre-built OBJs (which can be animated via plugin-driven datarefs) we can insulate plugins from drawing-environment changes.

Sunday, January 06, 2008

I Broke Volumetric Fog

I'll try to get it fixed soon....it's a bit frustrating, because it's the second time in the last week that I made a change to X-Plane to try to improve performance, tested it on my hardware, then discovered in-field that it helps some machines but screws up a lot of others.

The fog was a screw-up though...it's broken on the 9600XT and I have one of those.

(I'm not entirely sure what the minimum graphics card for volumetric fog will turn out to be...right now we let you use it no matter how slow it makes the computer, but generally it's been sort of a performance problem...it's just a very expensive algorithm that needs some kind of restructuring.)

Formats that are not changing

If you have content that worked in 864 but does not work in 9, please file a bug! Please do not assume we intentionally changed the file formats to break compatibility...I would rather get the bug report so we can fix X-Plane than have you change your work.

For every author who changes his or her work in response to a bug (an accidental breaking of compatibility), we'll have dozens of users who find their add-ons not working but not knowing how to fix it themselves.

With that in mind, beta 16 will have two fixes:
  • Some panels were showing the default instrument backgrounds over custom panels. This will also be fixed in the next beta.
  • Lit customized overlay parts were not showing. This will be fixed in the next beta.
So if you see those bugs, please check your plane against the next beta when it comes out. If you find other problems, please file a bug!

Saturday, January 05, 2008

Panel texture in weird places

X-Plane 8 didn't care much whether you used ATTR_cockpit in scenery objects or other strange places. It would simply show the cockpit panel texture, and if it hadn't been updated, you might see an old one, and if it had never been used, maybe you'd see the random (but colorful) contents of memory. Similarly if you could get close enough to another airplane to look in the window, you'd see your own panel, since there is only one panel texture (for the user's airplane) in the entire scenery system.

This is a bigger problem in X-Plane 9.
  • Because the panel texture can be expensive for big panels, we are a lot more aggressive about not setting up the panel texture if we can avoid it. This means that sometimes the texture doesn't exist at all. This is why in beta 14 you get an error if you do a formation flight having only been in "w" (forward 2-d) view...the panel texture doesn't yet exist, but the exterior view of the Cessna tow plane uses it.
  • With panel regions there can be up to four panel textures, so you can see the potential for anarchy.
  • Panel textures aren't even the same size any more, causing the wrong-panel-in-AI-plane problem to look even weirder than before.
So in beta 15, the panel texture is replaced with a dummy white texture for:
  • Any cockpit object for an AI plane.
  • Any scenery objects that are illegally using the panel texture.
This prevents crashes and other nasty stuff. If you want to make the panel be visible in your AI plane, consider using LOD to make a non-panel-texture "fake" cockpit image (at a very small res) at farther LODs. My guess is that in normal usage of the sim you'd really have to do something dangerous to get close enough to see the hack.

We did discuss live panels for all planes (for all of about 3 seconds), but the live panel texture in 3-d is so expensive that it'd be prohibitive to most users for even one AI airplane, let alone 20!

It's Time to Try Nine

If you create plugins, airplanes, or scenery for X-Plane and haven't tried your add-ons with X-Plane, please do so soon!! It's much easier for us to fix backward-compatibility problems while we're still in beta. Beta 14 introduced some bugs (that should be fixed in beta 15 real soon) but I think we're reaching the point where you can do compatibility testing.

We're working on a public Linux beta - see here.

Thursday, January 03, 2008

Beware the Forceware

Brett and a few other users pointed out to me that the nVidida ForceWare 169.21 release appears to hose video for X-Plane. If you have an nVidia card, don't update, or you may want to back up a version.

This kind of thing has happened in the past - hopefully a revision will come out fairly soon. (But I do not have contact with the nvidia driver team for Windows on this...)

Also some users are seeing corrupt startup screens with ATI hardware - apparently some configs don't react well to us turning vsync off. Not quite sure what we'll do about that yet, but hopefully a fix will be in the next few betas.

Is Bigger Always Better?

We've been preaching "one big texture, not lots of little textures" for a while now, and generally speaking, packing a lot of art into one big texture makes life eaiser for X-Plane, because it can draw more triangles at once before it has to tell the card to change what it's doing. Inside the company we call this the "crayon rule".

Now the total set of geometry and textures that X-Plane needs to use for one frame is the "working set" - you can think of it as the crayons that you keep out of the box because you need them all the time. And as I said before, if the working set becomes too big, your framerate dies.

Now with large panels we're seeing a new phenomenon, one of the first cases where the crayon rule might not be true. The reason is due to working set.

When you make an airplane with a large panel in version 9, you can either use ATTR_cockpit, which lets you use the entire panel as a texture, or you can use ATTR_cockpit_region, which will let you use several parts of the panel. Each ATTR_cockpit_region is a texture change, so that's more crayons. And yet ATTR_cockpit_region is usually faster.

The reason is two-fold:
  1. You can often use cockpit regions that don't cover the entire cockpit texture. Large panels are rounded up to 2048 if the are larger than 1024 in any dimension, so the "wasted space" in a 1600x1600 panel is actually quite huge. If you can get away with some smaller regions, your total panel texture area is smaller because there isn't wasted space due to this rounding, and you can also skip things like Windows. Prepping the panel texure takes time, and it's done once for lit and once for non-it elements, so it adds up!
  2. It turns out there are two categories of textures that contribute to the working set: static texures and dynamic ones, and their impact on VRAM is very different. Dynamic textures are much more expensive. The panel texture is dynamic and it's uncompressed, so it really costs a fortune. (32 MB of VRAM for 1600x1600. That's not a lot for a static texture but for a dynamic one that'll kill you.)
Here's the details on dynamic vs static textures: the OpenGL driver keeps a backup copy of a texture in main memory, so that if it has to purge VRAM (to make room for more stuff) it still has the texture. As it "swaps" textures, the process is to simply send textures as needed from main memory to VRAM. No big deal.

But with a dynamic texture, the texture has been modified in VRAM! So the copy in system memory is old and stale. The graphics card thus must send the texure back to main memory, consuming twice as much bus bandwidth as normal. (To free 16 MB of VRAM and refill it takes 32 MB of transfer, 16 MB to copy the old texture back to system RAM and another 16 to send the new textures to VRAM.) On non-PCIe cards, this back-transfer might be at 1/8th the speed of the transfer to the card, so this is even worse on AGP cards.

Thus the driver does its best to not throw out dynamic textures. And this is why the panel texture is so expensive. That P180 will cause X-Plane to make two 16-MB dynamic texures, and those textures will cause 32 MB of VRAM to basically be off the table. That's less space for the other textures to swap in and out of. This kind of "permanent allocation" makes the VRAM budget tighter for all other drawing operations.

Given the right combination of large panels, large res, pixel shader effects (which make more dynamic textures), clouds, and FSAA, you can easily get even a 256 MB card to a state where the free space into which static textures are shuffled becomes horribly small, and the framerate just dies.

So the moral of the story is: yes, it can be worth 4 crayons (using panel regions) to avoid the huge cost of dynamic textures from large panels.

As to static textures (regular DDS files) that are 2048x2048 - the jury is still out but my guess is they don't represent a huge performance problem. As one user pointed out to me, they're only 2 MB when compressed (maybe more with alpha) so they're not insanely huge, and they can be swapped out.