Direct X 10/11 Question - Answer (Translated)

[Cutriss]

3. DX10/11 are fluff. It's not a large step up from 9 in performance but it does have better GRAPHICAL capabilities, but then again this IS the generation of gamers who think graphics make the game.

Incorrect.

DX10's main draws are support for WDDM and pre-emptive multitasking (allowing multiple applications to use the GPU at the same time, which is what causes the current error when you alt-tab from fullscreen in XIV).

DX10 also draws a line in the sand for minimum capability and performance level, which means that devs using DX10 can be assured that the feature will work as intended on the target platform. No more muss and fuss about hardware T&L or embossed bump-mapping. It simplifies the code a lot and makes the experience more uniform. Sounds rather appropriate for something like Crystal Tools.

In short - don't talk about this unless you know what you're talking about, kthxbye.

[Betelgeuzah]

Yet what's the advantage of going DX10 if the current engine using DX9 is so horribly unoptimized?

If they said they were going DX10 that would upset me tbh. Optimize the game for the current direct X before aiming higher.

[Delsus]

Yet what's the advantage of going DX10 if the current engine using DX9 is so horribly unoptimized?

If they said they were going DX10 that would upset me tbh. Optimize the game for the current direct X before aiming higher.

Tbh if they upgrade now, any progress they have made on optimising the engine with Dx would not be compromised because the base code is basicly the same.

They could secretly be thinking of this, but dont wanna admit the core engine is poorly optimised for Dx

[Cutriss]

Yet what's the advantage of going DX10 if the current engine using DX9 is so horribly unoptimized?

If they said they were going DX10 that would upset me tbh. Optimize the game for the current direct X before aiming higher.

Remember, it's not just one game - it's all of Crystal Tools. If they're going to port any other games to the PC with it, then this is forward-preparation. And if not, then what was it all for?

[Betelgeuzah]

Remember, it's not just one game - it's all of Crystal Tools. If they're going to port any other games to the PC with it, then this is forward-preparation. And if not, then what was it all for?

I'd simply prefer if they prioritized. They probably don't have many programmers working on this thing, and I think at this point further optimizing the game for DX9 would have better results than going for DX10. Especially if they're not familiar with DX10.

[Jovy]

I figure if they are going to make major core changes to the game they might as well look at dx 10 as an option to start optimizing from scratch. I hate seeing my grafix card at 90% while my cpu is almost idle. Now if they feel they can optimize the game code in dx 9 and deliver a better in game expirience then i can respect that. Lets be honest, this game is seriously over budget and going to DX10 will only incurr more costs. Lets see what these summer updates bring us first b4 we bother SE to go to DX10.

[Cutriss]

I'd simply prefer if they prioritized. They probably don't have many programmers working on this thing, and I think at this point further optimizing the game for DX9 would have better results than going for DX10. Especially if they're not familiar with DX10.

That's the thing that baffles me about this the most. If it's that important...because they're going to use Crystal Tools for years, because they want the long-term income stream from XIV...why aren't they staffing the project?

Deus Ex ships in just over a month, that project uses DX11 and the Crystal Dynamics engine. If anyone can help them get this back on track, Eidos Montreal can.

[darkstarpoet1]

I'd simply prefer if they prioritized. They probably don't have many programmers working on this thing, and I think at this point further optimizing the game for DX9 would have better results than going for DX10. Especially if they're not familiar with DX10.

if they are not familiar with the up to date software then why not just get rid of them completely. they would be useless in trying to program games for the computer and pc. this is not a single player game that is expected to last a few months. a mmo is supposed to be good for years and to do this they should be able to handle what is out now that way they do not fall so far behind. all this does is lower the life expectancy of the game. if it starts behind it stays behind and becomes outdated faster.

[Ilean]

Jesus! This needs to be posted in every page of this thread:

1- Well optimized DX9 can look and perform great.
2- The game is on life support mode, even engine optimization as much as it's needed needs to be lower in priority than gameplay and content.
3- Crystal Tools isn't Yoshida's direct domain.

So again, DX10/11 isn't happening for this game any time soon, as much as I would love it, it just doesn't make any sense.

[Rentahamster]

Enabling DX11 IS optimization for those users who have the hardware that can support it. Stuff like ambient occlusion and cloth effects are able to be rendered with less strain on the system.

Take a case like WoW: Cataclysm, where you can run the game in DX11 mode. It looks exactly the same, but the game takes advantage of DX11 to more easily render certain things and boost your framerate: http://www.tomshardware.com/reviews/...ce,2793-7.html

In what almost seems like a stroke of irony, it looks like the folks most likely to see big speed boosts at the hands of Blizzard's DX11 code path are the ones with fast processors and near-overkill GPUs. If you built your PC with balance in mind, there shouldn't be as much room for API optimizations to augment performance beyond what you're already seeing. With that said, I've been passing this DirectX 11 flag around to a number of friends with lower-end Core i5 and Core 2 Quad configurations, and they're all seeing roughly 20 frame per second performance boosts. So long as your GPU is DirectX 11-capable, there's a fair chance that enabling Blizzard's code path will at least do something for your gaming experience.

It won't help everyone, but it's something.

What confuses me is that one of FF14's main selling points is the pretty graphics. The only games that are designed with DX9 only are old games, or games that don't have graphics as a priority.

Anyway, here is an example of how DX9 is denying the game of features -

A user on the JPN forums asked if we could have capes. This was the dev response (I don't believe it was ever translated here):

こんにちは。

FFXIVにマントを持ち込めるかどうか、開発チームと確認してきました。
残念ながらマントの実装は難しく、具体的な検討状態にもありませんでした。

やはり、マントを追加するからにはヒラヒラユラユラする必要があると考えていますが、今のFFXIVでヒラヒラユラユラさせるには、必要なスペックをグンと上げてしまうことになり、ちょっと現実的ではないという判断です。

Unfortunately, implementing capes is difficult, and we have no concrete plans for them currently.

Capes would of course need to be fluttering and waving, but to to so would require a large upgrade in the game specs and we decided that it would be impractical.

昨日の内容にちょっと言葉足らずの部分があったので補足です。

ヒラヒラユラユラするものが不可能です、というような強調した書き方になってしまいましたが、すべてのヒラヒラユラユラが無理ということでなく、例えばスカートは動きますよね。そういうものは大丈夫なんです。

マントで求められる、肩から腰を超えた長い装備を同様にヒラヒラユラユラするには、それを実現するためのモデルデータの都合上、難易度が非常に高いものとなる、ということでした。

Let me clarify what I meant:

It's not that fluttering and waving are impossible, as you can currently see such motion on equipment like skirts.

It's just that capes are large, spanning from the shoulder to the hips, and the model data required for rendering such a large area with fluttering and waving would be too high an increase to be practical.

FYI - it's easier to render simple cloth physics effects on DX11.