FSX-SP2(DX10): 3rd public post

[PhilTaylor]

I just posted another of my periodic updates on FSX progress with the DX10 update, here. http://blogs.msdn.com/ptaylor/archive/2 ... esday.aspx

I think you will find the progress interesting, and the results of the "DX10 looking like DX9" screenshots very interesting. At least I hope so :-).

[Nick N]

Links are working now!


DX10 http://blogs.msdn.com/photos/phil_taylo ... ginal.aspx

DX9 http://blogs.msdn.com/photos/phil_taylo ... ginal.aspx

[Nick N]

[quote]Thanks to both Phil and NickN for this update thread - and with its very exciting news of developments soon to come.

I am one of those, using Matrox triplehead 2 go, in Vista, with 4 gigs of RAM who just can't seem to get the out of memory problem fixed even applying the solution posted by Phil.

[Nick N]

I just posted another of my periodic updates on FSX progress with the DX10 update, here. http://blogs.msdn.com/ptaylor/archive/2 ... esday.aspx

I think you will find the progress interesting, and the results of the "DX10 looking like DX9" screenshots very interesting. At least I hope so :-).

Great news on the perf Phil, especially since that is without any optimizations


4 frames in FSX under the same scenery/conditions is actually a good boost and once opt

[born_2_fly]

Sorry if I'm being stupid.

But I see NO difference in the two pictures.

So much for the Artists impressions of DX10, with wave crests and more realistic everything, or was that a spoof?

I was hoping that I might upgrade my rig and move the FSX after the DX10 patch had been made, but I was pinning the idea of that on the basis that it would look even slightly better than FSX does at the moment.

If the only differences between the two is a performance increase. I hardly think that is work the money it would take to buy a DX10 card.

[PhilTaylor]

Sorry if I'm being stupid.

But I see NO difference in the two pictures.

So much for the Artists impressions of DX10, with wave crests and more realistic everything, or was that a spoof?

I was hoping that I might upgrade my rig and move the FSX after the DX10 patch had been made, but I was pinning the idea of that on the basis that it would look even slightly better than FSX does at the moment.

If the only differences between the two is a performance increase. I hardly think that is work the money it would take to buy a DX10 card.

Alex, did you take the time to read the post and think about it?

Read this again

That is the only difference you should see! Being done with the 3rd milestone, "DX10 looking like DX9", means all DX9 rendering features are now operational in the DX10 pipeline and look identical. That means the engine is fully flyable in DX10, but no new features are enabled.

there is supposed to be no difference in those shots, as they mark the completion of the "get all DX9 features working" phase. I even mentioned all 4 phases of the project to set context. In this post and the previous one. The 4th and final phase is "DX10 features" and I clearly say we are not done with that yet.

Is it clear now?

[Nick N]

I think many people are expecting to see some type of miracle visual improvement during a development phase.

The artist rendition images for FSX posted last year and other titles currently being developed in DX10 will tend to raise the bar on what is expected.

As Phil stated, this is a work in progress.

What is being shown is an accomplishment in the production schedule.

Because the team now has DX9 and DX10 looking exactly the same, and, the fact people are complaining about how much the images look the same, proves that milestone has been achieved.

Given the performance has increased in equal rendered environments without any optimizations in place also makes a statement to the rumors floating around that DX10 is worse than DX9 when it comes to performance.

There is still much to do, all the way up through DX10.1, for which you will need DX10.1 hardware to access, and I am assuming those features will also be coded into FSX as well.

[Fly2e]

Posted by: Nick N Posted on: Today at 12:28pm



I think many people are expecting to see some type of miracle visual improvement during a development phase.  

The artist rendition images for FSX posted last year and other titles currently being developed in DX10 will tend to raise the bar on what is expected.

As Phil stated, this is a work in progress.

What is being shown is an accomplishment in the production schedule.

Because the team now has DX9 and DX10 looking exactly the same, and, the fact people are complaining about how much the images look the same, proves that milestone has been achieved.  

Agreed Nick.
I think like you said people were expecting to see a side by side comparison that showed us the difference between DX9 and DX10 and what they are seeing is the same thing...... which is what Phil is saying.
This is a great accomplishment and as Phil said,

That is the only difference you should see! Being done with the 3rd milestone, "DX10 looking like DX9", means all DX9 rendering features are now operational in the DX10 pipeline and look identical. That means the engine is fully flyable in DX10, but no new features are enabled.  

Great to see all this happening and your updates are very much appreciated Phil!
Thanks again!

Dave

[born_2_fly]

Sorry if I'm being stupid.

But I see NO difference in the two pictures.

So much for the Artists impressions of DX10, with wave crests and more realistic everything, or was that a spoof?

I was hoping that I might upgrade my rig and move the FSX after the DX10 patch had been made, but I was pinning the idea of that on the basis that it would look even slightly better than FSX does at the moment.

If the only differences between the two is a performance increase. I hardly think that is work the money it would take to buy a DX10 card.

Alex, did you take the time to read the post and think about it?

Read this again

That is the only difference you should see! Being done with the 3rd milestone, "DX10 looking like DX9", means all DX9 rendering features are now operational in the DX10 pipeline and look identical. That means the engine is fully flyable in DX10, but no new features are enabled.

there is supposed to be no difference in those shots, as they mark the completion of the "get all DX9 features working" phase. I even mentioned all 4 phases of the project to set context. In this post and the previous one. The 4th and final phase is "DX10 features" and I clearly say we are not done with that yet.

Is it clear now?

Ahh, now I get you...

Sorry!  :-[

[MOUSY]

I found one difference. The DX9 render has more trees than the DX10. Does this mean anything and might there be any reason for that? Is it just because they're randomly generated that one may have more than the other? I was just scrutinizing the pics when I noticed this. I know autogen is just that - automatically generated - but might that slight difference contribute to part of the 4fps difference? (I know, certainly not all of it.)

So far, things are looking mighty good for the SP2. Everybody could use (and love) an FPS gain. 8-) Thanks for the update Phil.

[Nick N]

I found one difference. The DX9 render has more trees than the DX10. Does this mean anything and might there be any reason for that? Is it just because they're randomly generated that one may have more than the other? I was just scrutinizing the pics when I noticed this. I know autogen is just that - automatically generated - but might that slight difference contribute to part of the 4fps difference? (I know, certainly not all of it.)

So far, things are looking mighty good for the SP2. Everybody could use (and love) an FPS gain. 8-) Thanks for the update Phil.

I am at a loss to understand how you can count trees in those images and make any type of assessment that the amount of trees is reduced enough to even make a difference.

4 frames in FSX is like 6-10 in FS9 as far as I have seen in the way this title works.

Removal of all the trees in FSX SP1 gives me about 3-5 so the assumption that perhaps a few trees may be missing is making a 4 frame difference in the two images presented makes no sense.

Even if they are reduced in number, which again, I cant count a tree number even enlarging those images up 150%+. As far as I can see they are the same or close enough in count to where there could not be any skew of the results.

Let

[Ravang]

So does this mean I'll need to buy a DXT10 card or can I still use my DXT9 card with the update :-?

[Nick N]

So does this mean I'll need to buy a DXT10 card or can I still use my DXT9 card with the update :-?

Phil is better suited to answer this and can correct me if I am wrong, however in order to use DX10 and have its advantages you must have the following:

1. Windows Vista, which comes with DX10-10.1
2. A DX10 video card. Current cards are DX10, newer card cores will have DX10.1 ability

If you have a DX9 card you will still be flying fine but will not see or have any of the advantages DX10.x provides.


Unless I am mistaken, that

[JBaymore]

Alex, did you take the time to read the post and think about it?

Read this again

Phil,

Thanks for taking the time to clarify that for folks.  

I think the real issue is that while a lot of folks spend a lot of time USING computers...... they really don't understand what goes into the PROGRAMMING of computers ... and the complicated series of events, both technical as well as planning and scheduling, it takes to bring an app to completion.  Particularly at the OS type level.

I think a lot of people just seem to think this stuff somehow magically appears out of some "development tool" that does 99 percent of the real work.    Sort of like using "EZScenery".     (Aren't you guys just using EazyDX 2.0?)  

And reading these days is a skill that often is overlooked.

Thanks again,

.................john

[MOUSY]

[quote]I am at a loss to understand how you can count trees in those images and make any type of assessment that the amount of trees is reduced enough to even make a difference.

4 frames in FSX is like 6-10 in FS9 as far as I have seen in the way this title works.

Removal of all the trees in FSX SP1 gives me about 3-5 so the assumption that perhaps a few trees may be missing is making a 4 frame difference in the two images presented makes no sense.

Even if they are reduced in number, which again, I cant count a tree number even enlarging those images up 150%+. As far as I can see they are the same or close enough in count to where there could not be any skew of the results.

Let