Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Resolved [GSdx Hardware] Conflict - Desert Storm [PAL]
#1
***PCSX2 versions used:***
pcsx2-v1.3.1-1304-g059d4a1-windows-x86
pcsx2-v1.3.1-1292-gaff1467-windows-x86
pcsx2-v1.3.1-842-g1f402b1-windows-x86
pcsx2-v1.3.1-799-g000e11e-windows-x86
pcsx2-v1.3.1-699-gc660ea8-windows-x86
pcsx2-v1.3.1-641-gde644c5-windows-x86 (first available build with bug)
pcsx2-v1.3.1-590-gcd8e31d-windows-x86 (bug does not occur here)
pcsx2-v1.3.1-495-g4e82073-windows-x86 (bug does not occur here)
***Emulation settings:*** "Safest" preset, no hacks.
***Plugins used:*** GSdx-AVX2, cddvdGigaherz
***Plugin settings:*** CRC Hack Level "None"

#### Hardware Specifications 
***CPU:*** Intel® Core™ i5-4690K CPU @ 3.50GHz
***GPU:*** NVIDIA GeForce GTX 970 (Driver: NVIDIA 358.50)

#### Description
This is from the demo version in ops2m demo 33. 
None of the hardware renderers are able to draw the game without glitches.
Textures flicker, cycling from all black, wrong color. Even some HUD elements and text seem to flicker.
Software mode does not have the bug. Software mode without mipmapping also does not have the bug.

EE/FPU and VU0/VU1 Round/Clamping modes have no effect on the bug.
The bug also happens with the superVU dynarec.
None of the EE, FPU, VU, VIF1, DMAC, OPH Flag, Path 3 transfer related "game fixes" affect the bug.

#### Video
In the video I start with the OGL HW renderer, toggle back into the SW renderer and back again to HW.
https://vid.me/W0we

***GS dump*** 

.7z   ConflictDS-GS-FlickeringTexs.7z (Size: 2,74 MB / Downloads: 76)

Sponsored links

#2
hmmm , i wonder if this is the same engine as socom 2. Try some of the VIF hacks in the gamefixes section.

It's either that or path3 masking, the EE Timing fix may fix it or make it worse if that is the case.
[Image: ref_sig_anim.gif]
Like our Facebook Page and visit our Facebook Group!
#3
(10-21-2015, 11:41 PM)refraction Wrote: hmmm , i wonder if this is the same engine as socom 2.  Try some of the VIF hacks in the gamefixes section.

It's either that or path3 masking, the  EE Timing fix may fix it or make it worse if that is the case.

"EE timinghack", "Simulate VIF1 FIFO read ahead", "Delay VIF1 Stalls" and "Ignore Bus Direction on Path3 Transfer" are all turned on right now and the bug is still present. I'll repeat that Software mode doesn't have this bug, it's all in the hardware renderers.
#4
Oh doesn't it? Very weird then, must be a texture cache bug.
[Image: ref_sig_anim.gif]
Like our Facebook Page and visit our Facebook Group!
#5
I thought I managed to turn this bug off when I played this demo months ago. I'll have to look it up.
#6
(10-22-2015, 11:49 AM)FlatOut Wrote: I thought I managed to turn this bug off when I played this demo months ago. I'll have to look it up.

So there might be a months-old build that didn't have this bug, gotta track that down.

EDIT: I fired up the demo again, twiddled every knob with the latest build, no effect on the bug.
#7
could be worth trawling through old builds then to find out when it broke, it might give some indication of the problem, making it easier to fix
[Image: ref_sig_anim.gif]
Like our Facebook Page and visit our Facebook Group!
#8
Did more testing, I think I found the first build where the bug occurs. Can someone please validate it and push it to github?
#9
what build do you think introduced the bug?
[Image: gmYzFII.png]
[Image: dvedn3-5.png]
#10
(10-22-2015, 07:52 PM)Nobbs66 Wrote: what build do you think introduced the bug?

Sorry if I wasn't clear, everything is in post #1.




Users browsing this thread: 1 Guest(s)