FSX and DX10: Can it be fixed?

[Hiflyer 129]

A lot of us have experimented with the Dx10 preview mode of FSX and then gone straight back to the Dx9 Mode.

Even after drooling at the enhanced framerate that is available even with the special effects maxed, we have had to give a thumbs-down to the flickering runways and other texture problems.

And with the closing of Aces, it's been broadly assumed that these problems would never be fixed, but why is that, exactly?

Certainly there must be tons of talented programmers who are also flight enthusiasts who could have taken a look at the code to see if something could be done.......

Is it a legal issue?

At a time when so many old games are refurbished and brought back into the light with new, updated effects, it seems strange that the FSX community has not produced any developers/enthusiasts itching to tackle the problem.

Is it that hard? What might be involved, and what would need to be done to refurbish the dated FSX engine? Would it even be worth it to try?

I am just curious, as it seems like an idea that would have evolved naturally on its own, yet it has not, and I wonder why?

[mjrhealth 3]

DX10 was only ever implemented as a preview. of what possibilities it may have and even if aces where not closed down i doubt if anything would of changed until the next sim.

[Shaun Fletcher 570]

Hello Hiflyer,

I tried DX-10 whilst it may be more efficient for the computer the amount of distraction it caused due to ground textures flickering through on the runways etc made me not want to use it.

So to make the sim run more efficiently without it the hardware got upgraded

But I Think that FSX would have stayed at that stage and not been fixed either regarding the DX-10 issues..

[Thralni 24]

Why do you think that?

[Shaun Fletcher 570]

Why do you think that?

Hello Thralni,

Who thinks what sorry!

[Hiflyer 129]

Hello Hiflyer,

I tried DX-10 whilst it may be more efficient for the computer the amount of distraction it caused due to ground textures flickering through on the runways etc made me not want to use it.

So to make the sim run more efficiently without it the hardware got upgraded

But I Think that FSX would have stayed at that stage and not been fixed either regarding the DX-10 issues..

So you don't believe that the DX10 problems in FSX can be repaired?

[Shaun Fletcher 570]

Hello Hiflyer,

It probably could be fixed with some work, but what I meant was that I dont think it would have been fixed for FSX.

[Snave 466]

Be sensible. Just who would `fix it`..? FSX is hard-coded and the code is now locked away in the MS vaults, never to see the light of day. If it was easy to `fix` then two obvious questions should spring to mind, if you had thought for more than one second before asking the question:

1: If it was so easy to `fix` why didn't ACES fix it rather than release a last-minute hash-up of `preview` code...? Is it not patently obvious that they had some serious problems which proved impossible to solve?

2: If it was so easy for the aftermarket to fix, why hasn't it happened, more than two years after the patch that implemented `DX10 Preview` mode was released? Same insoluble difficulty or lack of access to the code. Or BOTH?

and, most tellingly of all (and this isn't a question, it's a FACT):

3: Why had ACES already publicly stated that it wouldn't be fixed even before they were told their jobs were gone..?

Final point to ponder: Who gives a monkeys cuss about `DX10`?

It's simply no longer relevant.

In little more than a month a DX11-based OS is on the market from MS. DX10 is old news and completely forgotten to all but us users of a completely abandoned product of a shut-down games studio.

[Hiflyer 129]

Be sensible. Just who would `fix it`..? FSX is hard-coded and the code is now locked away in the MS vaults, never to see the light of day. If it was easy to `fix` then two obvious questions should spring to mind, if you had thought for more than one second before asking the question:

Well, first we have to define what "sensible is." If you were a gamer then you might know that there are several old games (as I mentioned) that have been resurrected by fans and re-coded to more modern standards, so yes, I actually thought for several seconds before posing my question.

1: If it was so easy to `fix` why didn't ACES fix it rather than release a last-minute hash-up of `preview` code...? Is it not patently obvious that they had some serious problems which proved impossible to solve?

1) Please point to the part of the post where I mentioned that it was "easy" to fix.

2) Perhaps Aces were more focused on FS11 and wanted to save Dx10 for that?

3) Patently obvious that there were serious problems/impossible to solve? No, its not. Do you have any evidence to support that besides your own supposition? Did Aces or anybody else involved state such a thing?

2: If it was so easy for the aftermarket to fix, why hasn't it happened, more than two years after the patch that implemented `DX10 Preview` mode was released? Same insoluble difficulty or lack of access to the code. Or BOTH?

Again, please point to where I mentioned "easy" And as for why it has not happened, once DX10 was designated as a preview only for FSX, perhaps people just sat back to wait for FS11. Now that there is to be no FS11 and no further updates from MS, people might feel more free to go and dig around in the code.

and, most tellingly of all (and this isn't a question, it's a FACT):

3: Why had ACES already publicly stated that it wouldn't be fixed even before they were told their jobs were gone..?

Because they were focused on FS11?

Final point to ponder: Who gives a monkeys cuss about `DX10`?

It's simply no longer relevant. In little more than a month a DX11-based OS is on the market from MS. DX10 is old news and completely forgotten to all but us users of a completely abandoned product of a shut-down games studio.

Obviously to your thinking, its no longer relevant. But if we all had only one belief it would be a pretty boring world! I asked the question because I was curious about the answer. You have not answered, just told me why you think the question is irrelevant. Others may beg to differ. I would point out the many programs designed to update FSX textures to work correctly in DX10, and also the fact that REX (and other recent products) see fit to have DX10 settings available. Not to mention the many recent scenery's proudly attesting their ability to work correctly in DX10 mode. Obviously then, there are a few people who "Give a cuss" about DX10

I would also point out that many many years after Starlancer and Freespace were abandoned, there is a thriving modding community that has updated these ancient games with the latest DX10 effects and is still actively updating it. The same was done for quite some time for Falcon and other old Sims. Not to mention Starfleet: Bridge Commander, Oblivion, Crysis, and the list goes on. I simply wondered if the same might be possible for FSX.

[Snave 466]

Your simplistic notions are laudable, but the reason that I did not bother to reply further was that if you search through the blogs and websites of the former ACES team members you can find answers to the questions for yourself.

I do not provide `evidence` to back up my assertions is that it is rarely necessary to prove a negative, only to justify a positive and one never needs to back up a reply to painfully naive wish-fulfillment with factual justification, in the same way there really is precious little point in proving the Moon is NOT made of cream cheese...

There is no possibility of DX10 revamps for FSX for precisely the reasons I outlined: Whoever modifies it would need access to the raw code. NOBODY has it, and I don't need to justify that assertion when it's clearly been known for some considerable time. The other games you mention relied on tacit or overt involvement by the developers. Access to the base code or continued involvement by those responsible for the creation in the first place is a prerequisite.

NONE OF WHICH RELATES TO ACES/FSX.

You are also naive if you think that Microsoft would act in such a way after peremptorily cancelling not only the sim development, but also the game engine platform development that was based on FSX code (ESP).

My best insight is that FS* is such bastardised code after all this time, with patches, enhancements and a graphics engine that was creaking at the seams well before DX10 came along that the problems in invoking DX10 in a trouble-free manner were such that it was abandoned as a line of development in favour of beginning the process over - `FS11` if you will - and if ACES couldn't do it with the resources available to them and the will to succeed (at that time) to benefit Microsoft ESP, then the prospects are precisely zero.

DX10 is no longer a focus for development by graphics engine teams because Windows 7 introduces DX11 and 10, 11 or 2020 you would STILL be faced with the problem that graphics is NOT what limits FS, it's CPU-bound. No amount of tinkering with the graphics engine can resolve THAT, and without solving that restriction development on the graphic engine side is moot. AFAIK FSX is still a DX8-based engine, not even taking full advantage of DX9 features...

All of this is readily available information across the Worm Wild Wib and is there for the investigation if you want to research it for yourself.

Your best hope is for a new sim.

Our ONLY hope in fact.

[Hiflyer 129]

Your simplistic notions are laudable, but the reason that I did not bother to reply further was that if you search through the blogs and websites of the former ACES team members you can find answers to the questions for yourself.

I do not provide `evidence` to back up my assertions is that it is rarely necessary to prove a negative, only to justify a positive and one never needs to back up a reply to painfully naive wish-fulfillment with factual justification, in the same way there really is precious little point in proving the Moon is NOT made of cream cheese...

There is no possibility of DX10 revamps for FSX for precisely the reasons I outlined: Whoever modifies it would need access to the raw code. NOBODY has it, and I don't need to justify that assertion when it's clearly been known for some considerable time. The other games you mention relied on tacit or overt involvement by the developers. Access to the base code or continued involvement by those responsible for the creation in the first place is a prerequisite.

NONE OF WHICH RELATES TO ACES/FSX.

You are also naive if you think that Microsoft would act in such a way after peremptorily cancelling not only the sim development, but also the game engine platform development that was based on FSX code (ESP).

My best insight is that FS* is such bastardised code after all this time, with patches, enhancements and a graphics engine that was creaking at the seams well before DX10 came along that the problems in invoking DX10 in a trouble-free manner were such that it was abandoned as a line of development in favour of beginning the process over - `FS11` if you will - and if ACES couldn't do it with the resources available to them and the will to succeed (at that time) to benefit Microsoft ESP, then the prospects are precisely zero.

DX10 is no longer a focus for development by graphics engine teams because Windows 7 introduces DX11 and 10, 11 or 2020 you would STILL be faced with the problem that graphics is NOT what limits FS, it's CPU-bound. No amount of tinkering with the graphics engine can resolve THAT, and without solving that restriction development on the graphic engine side is moot. AFAIK FSX is still a DX8-based engine, not even taking full advantage of DX9 features...

All of this is readily available information across the Worm Wild Wib and is there for the investigation if you want to research it for yourself.

Your best hope is for a new sim.

Our ONLY hope in fact.

Ya' know what? I couldn't even make it through most of your post. You make a lot of definite assertions (quite a lot) then explain why it's not necessary to offer any proof of them, followed yet more assertions. There is an old saying about people saying things cannot be done generally being interrupted by people going ahead and doing it. We shall see.

[Shaun Fletcher 570]

Hello Folks,

Only time will tell in these instances

So lets all put 50p on it and leave it at that

[Aerosoft Team [Inactive Account] 51558]

Ya' know what? I couldn't even make it through most of your post. You make a lot of definite assertions (quite a lot) then explain why it's not necessary to offer any proof of them, followed yet more assertions. There is an old saying about people saying things cannot be done generally being interrupted by people going ahead and doing it. We shall see.

Well, I could follow what Snave wrote and all the information I got tells me that he is more or less right. That's information that comes from the ex developers (some of them are friends) and from current discussions with MS staff.

Of course MS 'could' make FSX work better with DX10 is they wanted to. But MS has never done a lot to keep their gaming products up to date (or even to fix open issues very efficiently. We got an internal list of bug in FS2004 that is many pages long and I can assure you that MS has a copy of that. But they never fixed those problems, not even when they were rather easy to solve. FSX got loads of patches in comparison to older versions that only had one small update.

There is no way MS will dump more money into FSX at this moment. They are 100% money driven in this market and they would never be able to make any money by making FSX better at this moment. To restart a stopped developments is costly beyond believe. It's a closed development, unfortunate for us.

[wedge 5]

hello.

I trust that if ACES team were still there they could have done something. But since they have been dismissed by Microsoft decision, it is very unlikely that any updates can be done to FSX. Although I'd like the people from ACES team to be able to work on their own and make more simulators with more freedom, I do think the process of creating such a piece of work is just huge, and it also requires lots of data, and many many things. But even if that were the case they could not do anything for FS, as it remains the property of Microsoft, the guys behind are not there anymore, but can't do anything for the Microsoft Franchise, well except working on add ons or stuffs like that.

I would also like to stress that even iof they were a new team, it would be if not impossible, very hard to go through the process of taking over what was done and managed by ACES Studio. It representes probably millions of code lines, and so many assets that it would be pretty hard. And to tell the thing even if there are tons of talented programmers in the world, a complex program like this is far from just programing, it implies many more things, like physics programing, 3D mathematics, and much more. I've tried to program some stuff and even a small thing like a 2D complete game represents already quite some work and time, but it stays easy, in 2D the maths are really not that complex. But entering the realm of 3D programming is really something different.

So well of course I would like some fixes to be done. But My hope is really for a new real complex simulator, which would surpass FSX, and have the same opening to add ons and extensibility, and keeping an intuitive interface. But FS is probably not coming back, or not in the way it was probably.

[Hiflyer 129]

Well, I could follow what Snave wrote and all the information I got tells me that he is more or less right. That's information that comes from the ex developers (some of them are friends) and from current discussions with MS staff.

Of course MS 'could' make FSX work better with DX10 is they wanted to. But MS has never done a lot to keep their gaming products up to date (or even to fix open issues very efficiently. We got an internal list of bug in FS2004 that is many pages long and I can assure you that MS has a copy of that. But they never fixed those problems, not even when they were rather easy to solve. FSX got loads of patches in comparison to older versions that only had one small update.

There is no way MS will dump more money into FSX at this moment. They are 100% money driven in this market and they would never be able to make any money by making FSX better at this moment. To restart a stopped developments is costly beyond believe. It's a closed development, unfortunate for us.

With my current computer, FSX just runs so smoothly! I just purchased 4 of the FsDreamtean airports, and when flying over them, the extra details and shading in the buildings (not to mention the lighting effects) are awesome. It makes me think that if fsx can display such things, that it was already so close to what it needed to be, and with a bit of love and care could have been around for many more years.

It seems almost a tragedy that the community will have to start all over again, practically from scratch.

[Snave 466]

I don't think you could be farther from the truth.

It's a `tragedy` that legacy code was allowed to go unrepaired and carried over from one sim version to the next, creating cumulative problems that were never solved.

It's a `tragedy` that ACES were forced to work to a common goal set that was never actually realised, due to circumstances outside their control, but very definitely within the control of the Corporation - and again these are well-documented, known and understood, so there's no need to go over old ground. FSX was supposed to have been the DX10 flagship product for Windows Vista, but it wasn't. And now never will be. Tragic? No. Disappointing? Yes.

The real benefit is that, with MS out of the picture, there is clearly a market for a freshly-developed product, free of the old necessities for retro-compatibility, terrible old legacy code forming the core of a product that goes back to DOS days, or even a need to accommodate ANY of the old models, practices or procedures of Flight Simulator: Why lament the passing when what it actually does is open the door for something else?

And there is great opportunity for creative thinking and suitable development, learning from the mistakes made by MS and ACES, and to offer something new that is far more than just prettier pictures - which is pretty much the only benefit you get from DX10 anyway, and arguably unnecessary when the extra details and shading can be achieved without it...

In any event, we are in a DX11 age now. Or at least we will be in about a month.

[Sharrow 107]

A lot of this "sudden" interest in DX10 & FSX is also driven by the performance increase that a lot of people are experiencing when using it. Mathijs' findings with regards to the 4850 and DX10 have also pushed more people to at least give it a try. Take a DX9 ManhattanX at 5 to 10fps and compare it to a DX10 ManhattanX at 15 to 20fps. The one is unuseable due to the low fps and the other is dodgy due to graphics anomolies... what a mess.

On the topic of DX11 will Aerosoft's new sim (if it goes ahead) actually be in a position to make any use of it? Any graphics engine under consideration at this stage would clearly be DX10. How would DX11 fit, if at all, into this picture in 12 or so months?

[BPL 39]

I must say that I am glad to be getting something new. MSFS was good, but not great. Let's hope Aerosoft's new sim will be great! I think it's refreshing to be starting from scratch. FSX is aging, and it's time for something new. The graphics in FSX are still good, but they're nothing spectacular anymore.

[Hiflyer 129]

A lot of this "sudden" interest in DX10 & FSX is also driven by the performance increase that a lot of people are experiencing when using it. Mathijs' findings with regards to the 4850 and DX10 have also pushed more people to at least give it a try. Take a DX9 ManhattanX at 5 to 10fps and compare it to a DX10 ManhattanX at 15 to 20fps. The one is unuseable due to the low fps and the other is dodgy due to graphics anomolies... what a mess.

On the topic of DX11 will Aerosoft's new sim (if it goes ahead) actually be in a position to make any use of it? Any graphics engine under consideration at this stage would clearly be DX10. How would DX11 fit, if at all, into this picture in 12 or so months?

And its all going to depend on your machine. As it is, ManhattanX is a solid and unchanging 20FPS at ultra high settings in DX9 for me. (except water, and no bloom, of course)

I am a fairly new enthusiast to the Sim, and in the last few months have spent some serious cash on planes and scenery that it would be nice to hold onto for a while. Right now, there is a thread at the REX site showing some examples of cloud shadows they are working on, and it looks great! It's the kind of innovation that makes one wonder just how far Fsx really could be taken, if people kept on pushing.

[Aerosoft Team [Inactive Account] 51558]

A lot of this "sudden" interest in DX10 & FSX is also driven by the performance increase that a lot of people are experiencing when using it. Mathijs' findings with regards to the 4850 and DX10 have also pushed more people to at least give it a try. Take a DX9 ManhattanX at 5 to 10fps and compare it to a DX10 ManhattanX at 15 to 20fps. The one is unuseable due to the low fps and the other is dodgy due to graphics anomolies... what a mess.

On the topic of DX11 will Aerosoft's new sim (if it goes ahead) actually be in a position to make any use of it? Any graphics engine under consideration at this stage would clearly be DX10. How would DX11 fit, if at all, into this picture in 12 or so months?

Personally I don't see a lot of graphics problems with most products in DX10. Or let me put it differently, I can live with them.

On our possible new sim, DX11 will almost be a necessity because we need to be ahead of the curve at the start.

[WEA-JHD 1]

I don't think you could be farther from the truth.

It's a `tragedy` that legacy code was allowed to go unrepaired and carried over from one sim version to the next, creating cumulative problems that were never solved.

It's a `tragedy` that ACES were forced to work to a common goal set that was never actually realised, due to circumstances outside their control, but very definitely within the control of the Corporation - and again these are well-documented, known and understood, so there's no need to go over old ground. FSX was supposed to have been the DX10 flagship product for Windows Vista, but it wasn't. And now never will be. Tragic? No. Disappointing? Yes.

The real benefit is that, with MS out of the picture, there is clearly a market for a freshly-developed product, free of the old necessities for retro-compatibility, terrible old legacy code forming the core of a product that goes back to DOS days, or even a need to accommodate ANY of the old models, practices or procedures of Flight Simulator: Why lament the passing when what it actually does is open the door for something else?

And there is great opportunity for creative thinking and suitable development, learning from the mistakes made by MS and ACES, and to offer something new that is far more than just prettier pictures - which is pretty much the only benefit you get from DX10 anyway, and arguably unnecessary when the extra details and shading can be achieved without it...

In any event, we are in a DX11 age now. Or at least we will be in about a month.

Very well said, Snave. You really hit the nail on this post. I agree with you 100%. It's also a reason why i think FSX was not a worthy replacement for FS9.

[Sharrow 107]

Personally I don't see a lot of graphics problems with most products in DX10. Or let me put it differently, I can live with them.

On our possible new sim, DX11 will almost be a necessity because we need to be ahead of the curve at the start.

I have heard "rumors" that you very seldomly fly the sim for fun anymore Mathijs (one of hazards of working with FS on a full time basis) so living with DX10 anomalies is perhaps a lot easier for you than for most?

Does this mean that the new sim (if confirmed) will use DX11 for sure?

[Sharrow 107]

And its all going to depend on your machine. As it is, ManhattanX is a solid and unchanging 20FPS at ultra high settings in DX9 for me. (except water, and no bloom, of course)

I am a fairly new enthusiast to the Sim, and in the last few months have spent some serious cash on planes and scenery that it would be nice to hold onto for a while. Right now, there is a thread at the REX site showing some examples of cloud shadows they are working on, and it looks great! It's the kind of innovation that makes one wonder just how far Fsx really could be taken, if people kept on pushing.

Indeed, though your 920 at 4GHz easily puts you in the top 5% of simmers out there...

The REX cloud shadow screenshots look fantastic btw! I just hope they can pull it off without grinding FSX to a halt.

[Hiflyer 129]

Indeed, though your 920 at 4GHz easily puts you in the top 5% of simmers out there...

The REX cloud shadow screenshots look fantastic btw! I just hope they can pull it off without grinding FSX to a halt.

The game is very sensative. Even the order in which you install things seems to have an effect, and so probably for the 5th time I am uninstalling and reinstalling everything very carefully now that I understand better which products make changes that in turn affect other products, etc not to mention which products replace things that are then falsely saved as original backups by following products....

I guess it will be the same with the REX clouds. (if they are ever released)

But the clouds are a good example of something that (apparently) was "impossible" being cleverly done anyway. With Orbx and some other packages I have seen, it seems some are definatly going to keep pushing things as far as they can.

[Snave 466]

The game is very sensative. Even the order in which you install things seems to have an effect, and so probably for the 5th time I am uninstalling and reinstalling everything very carefully now that I understand better which products make changes that in turn affect other products, etc not to mention which products replace things that are then falsely saved as original backups by following products....

I guess it will be the same with the REX clouds. (if they are ever released)

But the clouds are a good example of something that (apparently) was "impossible" being cleverly done anyway. With Orbx and some other packages I have seen, it seems some are definatly going to keep pushing things as far as they can.

Sorry, that is simply not so:

Cloud shadows were extant in early FSX beta test but were dropped in RTM for the simple reason that the frame rate hit was excessive. They were always technically possible under DX9... now, whether the patches to FSX and DX10 Preview compatibility messed that up subsequently I cannot say, but you are wrong to lump the cloud shadow issue with your fictional DX10 `solution`, in exactly the same way that the `popping` of autogen textures in SP2/Acceleration was a quite deliberate choice, not a fault.

But I do agree that taking responsibility for your own installation and understanding WHAT you are doing as well as HOW you are doing it is a major step for many simmers. The expression that is commonly used is `snake oil` and indeed MOST of the `Top Tips` and so-called `expert` advice have more in common with reptile juice than any basis in repeatable fact. You won't find many of those here as they tend to get debunked very quickly and most visitors here aren't the gullible type.

For that reason all simmers do need to have an understanding of the issues that FSX faces in the average system, and the only way to do that is not to repeat anecdotal evidence and rash assumption under the guise of a convinced argument or factual statement. And now you're doing it again. Perhaps a little more time learning, and a little less time opining would benefit your simming experience?

Doing so would allow you to see how little actual truth there is in your statement about installation order, but to understand the kernels of truth in what you have stated.