The New Faster, Leaner Next Generation Flight Simulator SIM-Posium is Now Available for Your Input

[cmpbellsjc 2,960]

My bet is that it will eventually fade into the darkness before something actually happens.

Someone will dig it up a few years from now and ask if anything has happened yet, or add a random remark, then it will fade back into the depths.

 

Maybe by then we will all be using Outerra or Dovetail will have their new Sim out by then.

[dtmicro 1]

Or maybe Orbx is secretly planning to take over the world by implanting a chip in everyone's head that says "just buy one more airport, one more"...

[cmpbellsjc 2,960]

Eventually someone will create a new SIM that surpasses what we have today, no idea how long it will be from now, maybe 2030, maybe next year

Yeah, eventually something will happen, but I doubt it will be from the main 5 to 10 people who have kept this thread alive this long. More likely a developer who already has something I'm the works and the funding to do so, with or without using talent from some 3rd world country who has top notch talent and will build it for pennies in the dollar.

Or maybe Orbx is secretly planning to take over the world by implanting a chip in everyone's head that says "just buy one more airport, one more"...

Judging from their user base, I think it may have already begun, lol.

[HiFlyer 9,343]

 

 

Yeah, eventually something will happen, but I doubt it will be from the main 5 to 10 people who have kept this thread alive this long. More likely a developer who already has something I'm the works and the funding to do so, with or without using talent from some 3rd world country who has top notch talent and will build it for pennies in the dollar.

 

This whole initiative was always a long shot. We've talked about this for years, and neither we nor the 3rd parties have acted. In the face of that fact, what options do we have? 

 

The answer is just as you said, a developer that already has something in the works, hopefully with something that finally takes full advantage of the power of our computers.

 

LM is really (to me) FSX all over again. X-plane is.... X-plane. And DTG? Who knows, but they are talking about Microsoft based technology, which might just turn out to be.... FSX with bells. (hope not)

 

So.

 

Who else out there is making anything with actual working software on the ground? It's been Outerra, and it's been my hope that we might be able to hook our wagon to their star.

 

Or we can get our crap together and go with Stephen B.

 

Or we can wait for DTG with crossed fingers.

 

It's our choice.

[dtmicro 1]

It's whomever gets the funding, it's their choice.

[cmpbellsjc 2,960]

This whole initiative was always a long shot. We've talked about this for years, and neither we nor the 3rd parties have acted. In the face of that fact, what options do we have?

 

The answer is just as you said, a developer that already has something in the works, hopefully with something that finally takes full advantage of the power of our computers.

 

Exactly, the community has tried via polls and petitions to MS to get them back in the flight sim business and it hasn't worked.

 

Then there was a guy who posted in the FSX forum a year or so back saying he was organizing something or was rewriting a new sim like FSX or something to that effect. That thread went a lot like this one and doesn't seem like anything ever can to fruition.

 

My guess is that either DTG, Outerra, X-Plane, or some other developer that has already been mentioned will do something.

 

You never know, maybe some of the existing FSX developers like ORBX already have plans to do something as a joint venture with DTG or some other big developer.

 

I just don't think Stephen B's website and all the back and forth here is going to spur anything.

 

In some cases all we can do is sit and wait.

 

It kind of reminds me of a thread I saw somewhere years ago where a fellow swore that he was going to organize a bunch of scientists to finally come up with a cure for cancer after the loss of a family member. As sad as it was that he lost someone due to cancer, all his wishing and rallying never brought any scientists out of the wood work to find a cure. He was reminded that their are already pharmaceutical companies,universities, independent teachers, etc. that have already been working on this for years and still have no cure and that's with funding. Of course nothing happened.

 

The whole series of events in that thread was much like this one.

 

While the OP's intentions are no doubt noble, I'm afraid that a we'll just have to sit and wait to see what unfolds from the folks who already have the money, talent and funds. Maybe I'm wrong, and Stephen will get something going, but I sincerely doubt it. Even if he does get something going a year from now, it will probably be a few years after that before something is ready to come to market, 2020.

[srborick 0]

Actually the idea of a SIM-Posium came about because I believed it was the flight simulator communities choice and that the community members would rally to create something on their own terms. 

 

If we don't make it our choice then we're stuck with the choices that are given to us by others. 

 

It began on the assumption that the community could pull together, discuss the possibilities, work out the logistics, etc. etc. 

 

This thread started May 26th, just 12 days ago and there have been over 16,000 hits. Yes I know that it's not 16,000 distinct individuals, but when you do the numbers it does tell me that curiosity has been sparked, and whether or not it's curiosity on how the debate is going or it's genuine curiosity in the possibilities of a next generation flight simulator I really don't know.

 

I do know that we will all have to live with what is handed to us unless we take all of the talent that is available within the community and create that plan that was talked about on this thread, it will take all of us together to determine the right tools to use, and it will take all of us to pitch in financially if we really want to have a choice.

 

I am amazed when I look over this thread and read the actual discussions about flight simulation and the building of one, the cost factors, the 3d engines, the idea of multiple sims vs a single sim, etc. etc. All of that was done in just 12 days.

 

Now... try to imagine what could happen in 180 days if more and more talented individuals joined in the discussion.

 

It is possible, it can be our choice... whether or not we choose to make it our choice is up to us.

 

Stephen B. 

[dtmicro 1]

Actually if we can use Outerra to create the world, and then Outerra can help us synchronize and overlay some custom Terragen images into their engine, then I think we have the best of everything.

 

WE start with a procedurally rendered global world, and then use Terragen to enhance it.

 

I actually appreciate the discussion and didn't meant to sound demeaning to anyone at all. No-one knows everything, I was just sharing what I thought I knew about it, which isn't as much as some do. Just my thought.

This is an example of my work in the Utah Area, starting from very low-quality 1.2m NAIP images, but restored using tons of advanced imagery tricks in Image Magick and other apps:
 

Desert:

http://www.mediafire.com/view/i2evx5t5632e8xl/SceneryFX_VastUtahDesert.jpg#

 

Mountains:

http://www.mediafire.com/view/66h3uemsje7keoo/UtahMountains.png#

 

Sorry posted wrong link accidentally...

 

The problem is running it through that process greatly increases the time of the photo correction and blending between tiles where errors exist from color corrected zones.

 

Another issue is that this is 1.2m... Will post a 60cm example...

 

It just takes ME SO SO long to do this as one person, that's all.

[cameni 9]

That was not the first time I read about Outerra, I've read about it many times. I'm also aware of other programs like it. I only just tried the demo the first time, as I said it shows promise, but procedural generation of textures has limitations in the variations you can achieve as well as to try to tweak things to look like the real world.

 

It would be difficult relying purely on math adjustments to get textures to look right, unless the engine were highly refined. It would also not be as add-on friendly.

 

Which is why if I were to start today, I'd be making the textures in Terragen, not until "on-the-fly" procedural generation can impress me as good as Terragen pre-rendered images can.

Terragen is an offline procedural renderer. It's a great procedural renderer, but it's unusable for real-time rendering, as it was build with different objectives.

The only way you can use it for games is to pre-render all textures and meshes and store them somewhere for the engine to access. Also, it means the engine now has to be optimized for streaming, and even though people easily dismiss FSX engine as outdated, it was heavily optimized for streaming too, and not using the GPU so much because of the same architecture type.

 

Outerra is a real-time procedural renderer, moving the load to the GPU. Obviously it cannot reach the same quality as Terragen in 1/30 sec on today's hardware, it has to do a lot of compromises. It's also alpha, and the variety it can generate is currently limited because of that as well.

 

But to put in into a perspective: just the land textures generated by Outerra in run-time (2x2cm² medium res) would take 1.3 exabyte (1.3x10¹⁸) of storage on the disk for the whole planet Earth. Actually you would not be able to use them because suddenly the throughput from the storage or net to the GPU will become a big issue.

Time it will take Terragen to produce the data will be huge as well. And you won't reach anywhere nearly the same quality as movies produced by Terragen frame by frame ...

 

You may disregard the need for such textures in a flight sim if your objective for the next gen sim isn't a world simulator, but you are still keeping all the disadvantages of a streaming engine, you just changed how the imagery and meshes are produced.

 

As a side note: Outerra was inspired by Terragen 1, with the goal of making the procedural generation real-time and usable for games/sims.

[dtmicro 1]

No matter how good they get at doing the above method and creating content and textures with equations on-the-fly, there will always be a procedurally generated flavor to the images, with pre-renders you can create any flavor without having to make code modifications that add further alteration parameters to the procedural equations.

 

--------------------------------------------------------------------------

Why are you making the same mistake as the other guy did and trying to re-explain what we already went over so many times, it's going in a circle. We know Outerra is a real-time "on-the-fly" procedural terra-generator from the very beginning, and I also stated Terragen would take up too much HD space. This was never in question, never...

 

I am talking about enhancing in the sense of overlaying, not editing the images. That way you don't need as much Terragen rendered content as you would if you used Terragen only, some procedural some rendered. The issue with using Terragen images is they have to be down-sampled and MIP MAP'd, yah well there are tools to do that already, but you do have to set some parameters correctly. So even after that process

 

I know what Terragen and Outtera is and the difference, my rendering experience goes back to the 1980's

I've used 3DS Max, Cinema Lightwave (whatever it was), Terragen, Vue 'D Esprit, Maya, Mental Ray, to name a few...

 

I come from the Amiga and C-64 days, heck even before that.

 

--------------------------------------------------------------------------------------------------------------------------------------------

See the issue is with those 2 images I posted above, that is about the limits of how good you can make NAIP 1.2m imagery look. Sure for some places you can get 30cm data or 60cm, but still more on the East Coast for the US than the Western Rockies. You can only make them look SO good, and at some point you just really want to render the darn things instead of using the Photo-Real Scenery to correct them :_)

 

That's why Orbx always releases pure PR scenery in such small batches, the correction time is so drawn out sometimes.

[dtmicro 1]

The amount of space of course, depends on the detail of the base renders, and whether using landclass like repeat textures vs. fully custom real (instead of PR though, we are now using Terragen ahead of time on top of the Outerra). It's basically the same concept...

 

My problem with Outerra is it will NOT look like the REAL area in real life (it might look somewhat like it), and it will get boring fast just like FSX, it might seem cool for a few minutes, but it will get very repetitive, regardless of how many different ways they can make the equations produce the graphics to look different for different areas.

 

That said, I think it is plenty good to use it as a global base for textures.

 

Also, it means the engine now has to be optimized for streaming, and even though people easily dismiss FSX engine as outdated, it was heavily optimized for streaming too, and not using the GPU so much because of the same architecture type.

 

The main reason the FSX engine is slower to load than something like a GTA V engine is 4 reasons:

 

1) FSX Caching method and memory style, GTA smaller area

2) FSX MIP MAPs are too large, or people didn't even MIP MAP some things...

3) FSX does not auto-gaussian everything as much as it should

4) There are other things it didn't do as well that are newer on more modern video cards

 

As I noted earlier, if you play GTA V and fly at some distance from their mountains, you will quickly see a GAUSSIAN BLURRING trick kick in to reduce the texture memory load on the Video Card.

The reason Rockstar is good at that trick and have mastered it is because they are CONSOLE developers, and on the consoles that trick is required even more so than PC's usually or you'd get crashed... All the big Console developers that have made massive game worlds have used some form of that trick (fog or blurring, or both).

 

Also, for Flying I think it has already been assumed people want DETAILED AREAS, not just one mass global templated equation maker area.

Maybe for now they care a little, but it will be common to have 8+ TB drives by the time this PROJECT would be released. Also SSD's will be cheaper, so most will have 1 huge non-SSD drive and a smaller 2TB to 4TB SSD drive by then.

 

Pre-Rendered vs. Procedural On-The-Fly

Pre-Rendered content can greatly spice up procedurally generated content. Engines like Outerra type generators are nothing new, it's just that Outerra has done it in a way that is acceptably nicer looking due to following Terragen style renders. We've had procedural fractal generation programs since the 1980's, it's just back then they were mostly almost like a WIRE MESH generator instead of having the added ability to generate textures with similar equations.
 

You can make any Outerra like image with Terragen just as easily by rendering it ahead of time, the advantage of having procedurally generated content is only because it saves hard drive space and can potentially load larger areas faster, so not only can it GENERATE graphics from scratch from equations, but the working of those equations in the Memory Address Space, you can refer to the same address space over and over and over again, or you use mathematics and shaders and warp functions to slightly alter the object and then keep referring to the base object by inheriting it. Yes basic programming concepts do sometimes apply in graphics too.
 

Also some games even way back in the 1990's used to procedurally randomize (only in 2D though) with top-down perspectives a random world to play in, with random mountains (forgot name of the games, been so long ago, would have to think about it).

[cameni 9]

I'm assuming you are not a game/engine programmer, just a long time user of non-realtime rendering software, right? Not meaning to be derogatory at all, just wondering, based on the choice of words. Actually, I find the usage of terms very weird. What does it even mean that texture mip-maps are too large? Auto-gaussian ... refers to GTA engine basically hitting limits of their supported level of detail range.
And so on ... you are even assuming that Terragen renders = engine renders of Terragen data.
 
Yes I'm assuming that people will still care where to put exabytes (million TB) of data on their next gen 10TB disks.
 
This problematics of large scale high fidelity world rendering goes so much deeper than just wishfully combining pieces of software that one likes, as if it could somehow auto-scale beyond the normal annual hardware performance boost.
 

You 2 should go around and around in circles using the PM system so we don't have to watch you both chase your tails.

I fear I made a mistake trying to explain and collapse that circle :huh:

 

[dtmicro 1]

No one is talking about rendering the entire world using non Land-Class Terragen to render every detail, and sorry if you think some of my terminology is weird.

 

-------------------------------------------------------------------------------------------------------

Hate to disappoint you, but the best results in games we've seen have not been from procedural generation, and GTA V is not really all that procedural, some parts of it are like roads over certain areas, but the majority of custom TEXTURES are land-class type method (if comparing to FSX, repeat textures painted over a mesh). Creating textures procedurally for the whole world and leaving it there, well obviously you are going to EXTEND that. The textures were (Drum roll) based on Photo-Real but also some were rendered over the mesh. They are using exactly the same technique I was discussing before. Just Cause 2 was more procedural, but it was still a mix, but yes that did make Just Cause 2 get boring fast and, even though the graphics were good. And I wouldn't say Outerra in its current state better than Just Cause 2 really (but Outerra has potential as a base).

 

We will see how Just Cause 3 did with the procedural re-pasting of stuff, or if they improved the technique.

 

Yes, you need the add-on developers to be able to enhance scenery of the map without having to use sliders for the adjustments or shaders of an area and instead of creating some type of procedurally generated texturing over teh area (I doubt Sim developers want to do it that way). To try to make it look more like the real thing, you need to use other tools.

 

My terminology is not weird, too large or not enough MIP Maps, really?

It means there are NOT enough MIP MAPs in the radius, or in some games that the MIP MAP was too big in size, and it is taking up too much memory and low VAS. It is a big problem in FSX with some developers either not creating the MIP Maps, but even more so the FSX engine not being able to unload some of the textures by blurred MIP Map type things.
 

This is why FSX can actually have sharper images than GTA V in mountains if you are far enough away when using REALLY high res imagery, the MIP MAP and blurring isn't enough in FSX.

[cameni 9]

No, I never heard the term "mipmap too large" to mean there's not enough coarse terrain texture detail levels - if that's what you mean.

 

As for the rest, you aren't arguing with me, please do not reinterpret what others say. Procedural is used to fill up the world with detail that would be impossible to produce manually at that scale. But procedural is also the application of vector data in engines that can support vector overlays. Designers then can combine raster and vector data for custom modeling of selected areas.

[dtmicro 1]

Since you have never heard the term of a MIP MAP too large, then I suggest you read the following PDF,
Computer Graphics: Principles and Practice,
 

Specifically page 494 which I linked to below, because it uses that terminology.

https://books.google.com/books?id=OVpsAQAAQBAJ&pg=PA494&lpg=PA494&dq=mip+map+too+large&source=bl&ots=QYKgohn5UU&sig=Ei_XEtjiuTr3AzV2PX-OrMMcyPc&hl=en&sa=X&ei=_iB0VZLeMcfloASXz4DYDQ&ved=0CDQQ6AEwAw#v=onepage&q=mip%20map%20too%20large&f=false

 

It doesn't mean not enough coarse terrain, that's not what I said.

It means it's taking up too much memory given its relative nature to the rest of the graphics in the radius, or that the radius is in danger of consuming too much memory, or that the MIP MAP'd object(s) actually will not fit into the memory that is left over.

 

It seems one technique to commonly help this is by actually gaussian blurring to hide the quality of the MIP MAPs so you can then load the highest LOD in a smaller radius ("closer to the viewer").

 

I'm not an expert at how that's done by various engines because I've never written a game engine (but yes I have written some gaming stuff), but I see it all the time, especially exaggerated in Console games as it is often used severely in this case (been seeing it happen for years and years), since Consoles generally have lower specs.