Quote:
Originally Posted by Heliocon
Well yes and no. The reason the clouds and water are behind is because the tech that would allow them to really push clouds and water graphics are dx10/dx11. Dx11 for direct compute (physics), and differed lighting along with other lighting and particle effects that are in dx10 (but much improved in dx11), which could greatly reduce the burden on rendering these elements.
DX11 is not a niche market, there was a huge thread/argument about this last year and I made a prediction back then about market share and cards, which was that by 2011 win7 64bit would be the primary OS and that dx11 card market share would skyrocket at the end of 2010 and for the first half of 2011. Months after and even now, steam 100% backs this up and at the current rate in 6 months dx11 may even be the majority (by % increase last month out of the 15 cards that had the strongest % growth in market share, 14 of them were DX11 and represent probably around roughly 8%+ growth for dx11 card market share last month alone).
DX11 isnt only a rendering path, it has a lot more to it which makes it far better then dx10, alot of that is it is made to work in sync with multicore cpus/multi threaded apps. It has a new shader model which gives alot of new features (shader model 5, dx9 = shader model 2).
Also crysis is consolised to hell and back, thats why it has no dx11 on release, in its video settings there are 3 options, gamer - advanced - hardcore and there are literally no descriptions of even what each setting has in it... The tutorial uses console gamepad button instructions etc etc. Also the engine looks worse then the original crysis did.
|
We are in the first half of 2011, the game is releasing in 2 weeks, with DX11 market share still <6%. So if Maddox really pushed for DX11, they are sacrificing their time for that 6% instead of the remaining 94%. This is the same exact reason why Crysis 2 got scaled back, they got burned for pushing too hard on the first one. I'd have to disagree with you on Crysis 2 looking worse than the first one, but let's at least agree to disagree on that one.
Also, how will DX11 sky rocket? The rate of change is more or less linear, from 0-5%. If anything, history indicates most changes are linear, not exponential according to the steam graphs.
Sandy Bridge is not DX11, Llano won't be out till end of this year, Ivy bridge next year - all the mainstream DX11 parts will need time for uptake too.
Also, I have not heard of a single game released/being released that's heavily based on DirectCompute. The development started 5 years ago, reality will not have allowed this to happen.
What's important is that the engine is built to be scalable - that's what future proofing is, and if that's the case, I'd agree with Maddox' assessment that this is THE next-gen engine. DirectCompute, tessalation, shaders can be added when the return of investment seems solid.
Having DX11 currently on CloD will not make it better - from the released screenshots and videos, this game suffers from polish and optimization. Having one more item on the list will further deteriorate the quality of this game.
Unfortunately we have to accept the realities of our flight sim niche market. We don't have 300 men team working on a single game. Beggars can't be choosers.