Reply 121 of 389, by lowenz
Reply 122 of 389, by lowenz
With SwiftShader ( https://swiftshader.googlesource.com/SwiftShader / https://drive.google.com/drive/folders/0B8_Bl … cEhTdy1UNzhXSVE ) the image is perfect, but @2 FPS :p
Reply 123 of 389, by Oubadah
wrote:Look at the MISSING water transparency with WineD3D 1.9.21.....
I wasn't missing anything with 1.9.21:
Reply 124 of 389, by lowenz
The *water* transparency.
Time to test with 2.2 staging.
Reply 125 of 389, by Oubadah
wrote:The *water* transparency.
Time to test with 2.2 staging.
My screenshot shows the same area of water that yours does. If you can't see the whole image, right click and open in new tab/window. This site can't properly reformat big images evidently.
Reply 126 of 389, by lowenz
Well, now the problem is gone with 2.2
Reply 127 of 389, by Oubadah
There was something weird going on it those screenshots of yours anyway. Where were the tree and rock reflections?
Reply 128 of 389, by teleguy
wrote:There was something weird going on it those screenshots of yours anyway. Where were the tree and rock reflections?
He probably has r_Quality_Reflection set to 0.
Reply 129 of 389, by Res1s7
Hi. Has there been any update to this ? Is it a frame rate killer ? Or maybe I just had it setup badly last time. GOG: Should Far Cry bought from GOG have this problem fixed ? Is it responsibility of GOG ? I know it can't always be done. Original IP owner has to want to fix it. Availability of original source code, etc, etc.
Reply 130 of 389, by lowenz
First Binocular Point save: 290 FPS vs 120 FPS (i3 6300 + GeForce 1050 Ti) using Wine 3.0 RC1
Wine always necessary to fix the landmass reflections with the Steam version.
Reply 131 of 389, by CookiePLMonster
- Rank
- Newbie
Hey there,
Looks like WineD3D is still the only solution to fix water reflections and I never really was satisfied with this solution, so decided to give it a try and... there it is!
Water reflections rely heavily on user clip planes (so bottom surfaces of water pools don't get reflected), and for reasons unknown behaviour of IDirect3DDevice9::SetClipPlane has changed between XP and Vista/7/8.1/10. Even though D3D9 is supposed to work like a finite state machine, state of clip planes did not persist properly between draw calls (either that or maybe there is some precomputation applied to it) - so my fix is to re-apply clip planes before every draw call.
Works flawlessly, native D3D9, no performance hit!
For reference that it's legit (though I admit the timing is sub-par), I'm an author of several unofficlal patches for other games.
Cheers!
Reply 132 of 389, by lowenz
Thanks man!
Reply 133 of 389, by Johnylyr
wrote:Hey there, […]
Hey there,
Looks like WineD3D is still the only solution to fix water reflections and I never really was satisfied with this solution, so decided to give it a try and... there it is!
Water reflections rely heavily on user clip planes (so bottom surfaces of water pools don't get reflected), and for reasons unknown behaviour of IDirect3DDevice9::SetClipPlane has changed between XP and Vista/7/8.1/10. Even though D3D9 is supposed to work like a finite state machine, state of clip planes did not persist properly between draw calls (either that or maybe there is some precomputation applied to it) - so my fix is to re-apply clip planes before every draw call.
Works flawlessly, native D3D9, no performance hit!
For reference that it's legit (though I admit the timing is sub-par), I'm an author of several unofficlal patches for other games.
Cheers!
So where is the patch? And what about d3d10;
Reply 134 of 389, by CookiePLMonster
- Rank
- Newbie
wrote:So where is the patch? And what about d3d10;
Soon! And not sure what you mean by D3D10, as far as I'm aware FC1 doesn't have a D3D10 renderer.
Actually, is there anyone with AMD (or Intel) GPUs here who could volunteer as a tester for this? So far I could only confirm it working on NVidia GPUs.
Reply 135 of 389, by Srandista
- Rank
- Oldbie
wrote:Actually, is there anyone with AMD (or Intel) GPUs here who could volunteer as a tester for this?
I have both of them, if you want to test it on Win 10.
Socket 775 - ASRock 4CoreDual-VSTA, Pentium E6500K, 4GB RAM, Radeon 9800XT, ESS Solo-1, Win 98/XP
Socket A - Chaintech CT-7AIA, AMD Athlon XP 2400+, 1GB RAM, Radeon 9600XT, ESS ES1869F, Win 98
Reply 136 of 389, by lowenz
The patch will be a modified D3D9.dll? I ask 'cause of reshade......
Reply 137 of 389, by CookiePLMonster
- Rank
- Newbie
wrote:wrote:Actually, is there anyone with AMD (or Intel) GPUs here who could volunteer as a tester for this?
I have both of them, if you want to test it on Win 10.
Cool! I'll post when I have something which can be tested.
wrote:The patch will be a modified D3D9.dll? I ask 'cause of reshade......
That's what I was worried about... I did want it to be d3d9.dll (because the game breaks Ultimate ASI Loader badly at the moment), but you have a good point... I'll see if I can make it into an ASI plugin. Will most likely need to ship ASI Loader as gdi32.dll then, as pretty much no other DLL works well (dinput8 loads too late and dsound is unreliable).
However, I may also publish this as d3d9.dll and issue a pull request to ReShade with the same changes.
EDIT:
Interesting - looks like the game supports a D3D8 renderer! Did any version of the game come with XRenderD3D8.dll?
Reply 138 of 389, by lowenz
Reply 139 of 389, by lowenz
wrote:dsound is unreliable
And taken for Indirect Sound 😁 -> http://www.indirectsound.com/
However, I may also publish this as d3d9.dll and issue a pull request to ReShade with the same changes.
Good idea!