mirror of
https://github.com/hrydgard/ppsspp.git
synced 2026-04-25 06:06:03 +03:00
[GH-ISSUE #5001] DTM Race Driver 3; Toca 2 Graphic Problems #2095
Labels
No labels
Atrac3+
Audio
CPU emulation
D3D11
D3D9 (removed)
Depth / Z
Feature Request
Font Atlas
GE emulation
Guardband / Range Culling
HLE/Kernel
I/O
Input/Controller
MP3
Multithreading
Needs hardware testing
Networking/adhoc/infrastructure
No Feedback / Outdated?
OpenGL
PGF / sceFont
PSMF / MPEG
Platform-specific (Android)
Platform-specific (Windows)
Platform-specific (iOS)
PowerVR GPU
SDL2
Saving issue
User Interface
Vulkan
arm64jit
armjit
armv6
x86jit
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
starred/ppsspp#2095
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @zschigi2 on GitHub (Jan 4, 2014).
Original GitHub issue: https://github.com/hrydgard/ppsspp/issues/5001
in DTM Race Driver 3 and Toca 2 there are massive Graphic Problems ingame
(some kind of scan lines over the game-graphics)
@dbz400 commented on GitHub (Jan 4, 2014):
Can you post the screenshot ?
@zschigi2 commented on GitHub (Jan 4, 2014):
yes :-)
@dbz400 commented on GitHub (Jan 4, 2014):
How about in software rendering mode ?
@zschigi2 commented on GitHub (Jan 4, 2014):
in software rendering mode i only get a black screen...
@dbz400 commented on GitHub (Jan 4, 2014):
Okay i can reproduce it now in both GLES and softgpu
@unknownbrackets commented on GitHub (Jun 9, 2014):
Has this improved with simulate block transfers?
-[Unknown]
@zschigi2 commented on GitHub (Jun 9, 2014):
unfortunately it does not. the problems are still the same. (tested on my s3)
@unknownbrackets commented on GitHub (Jun 9, 2014):
Does it have the same problems on Windows? Maybe it's an armjit bug. Would help to narrow it down.
-[Unknown]
@zschigi2 commented on GitHub (Jun 9, 2014):
it´s the same on windows... :-(
@unknownbrackets commented on GitHub (Jul 20, 2014):
Can you try to find the problem on Windows using the GE debugger?
https://github.com/hrydgard/ppsspp/wiki/How-to-find-a-graphic-issue-with-the-GE-debugger
-[Unknown]
@zschigi2 commented on GitHub (Jul 20, 2014):
if i follow the instructions i get these:
@unknownbrackets commented on GitHub (Jul 20, 2014):
Well, the goal is to find it when the GE debugger is drawing the thing that shows up wrong. In your screenshot, it seems to be drawing some sliver in a separate framebuffer... this is interesting (btw, you are using buffered rendering, right?), but it doesn't really show what's causing this graphical issue.
Keep pressing step prim until you start to see the correct graphics. It highlights in red the part it's about to draw (as in your screenshot.)
-[Unknown]
@zschigi2 commented on GitHub (Jul 20, 2014):
this time with buffered rendering on:
@unknownbrackets commented on GitHub (Jul 20, 2014):
Okay, but those are just it drawing the screen still, right?
It's gonna spend a lot of time doing that, you can try pressing "Step Tex" if you're impatient, but if you pass where it draws the graphical glitches you have to press "Step Frame" and start over.
-[Unknown]
@zschigi2 commented on GitHub (Jul 20, 2014):
here we go...
@ppmeis commented on GitHub (Jul 25, 2015):
Tested in latest build. Same status:
@ppmeis commented on GitHub (Dec 6, 2015):
Tested with latest build. Same status.
@unknownbrackets commented on GitHub (Dec 21, 2015):
Are you able to capture when the GE debugger's red overlaps with one of the red lines specifically (e.g. not just the full screen, but is only over the black line, or right at the black line?)
The earlier screenshot looks like a clear and is probably not related.
-[Unknown]
@hrydgard commented on GitHub (Apr 1, 2017):
Probably broken geometry that gets discarded by the real PSP because the coordinates are way outside the guardband, and the PSP doesn't do proper triangle clipping. I'll send a pull request later with an experimental fix for D3D11 but it relies on undefined behaviour in vertex shaders to kill triangles, which isn't very nice. A "correct" fix will need geometry shaders.
@unknownbrackets commented on GitHub (Sep 22, 2018):
Hmm, reported that DTM is not fixed in #11393.
A GE dump would be useful.
-[Unknown]
@LunaMoo commented on GitHub (Sep 23, 2018):
Works fine on windows at least, so that report might be another one from android where it might not had enough vertex registers or that's how the "optimization" fixing other games mentioned there is working aka it might be completely disabling the guardband fix.;]
@hrydgard commented on GitHub (Sep 23, 2018):
Or forcing the coordinate to NaN isn't doing the right thing on the hardware....
@marosis commented on GitHub (Sep 28, 2018):
It is working right on every backend? I tested it on OpenGL on android and it is not working for me
@marosis commented on GitHub (Sep 28, 2018):
@marosis commented on GitHub (Sep 28, 2018):
We need more testers
@hrydgard commented on GitHub (Sep 28, 2018):
@unknownbrackets maybe it would help to set all four parts of the position vector to NaN, not just w...
@marosis commented on GitHub (Sep 28, 2018):
Now i tested on latest build if i didn't get Wrong info and yeah STILL SAME status.
@hrydgard commented on GitHub (Sep 28, 2018):
Thanks marosis. Stay tuned, I'm going to try my idea in the weekend.
@marosis commented on GitHub (Sep 28, 2018):
I will
@unknownbrackets commented on GitHub (Sep 28, 2018):
That sounds worth trying. The cull distance extension isn't supported by much other than Adreno, so that probably won't help.
Maybe the other option is to try to set depth to a value that would be (mostly?) clipped and then also discard in the frag shader, but that sounds pretty awful...
-[Unknown]
@ghost commented on GitHub (Sep 28, 2018):
This is the only graphics glitch I encounterd in my Mali-450 MP I think it's normal for my GPU same graphics issue in Monster Hunter 3rd HD
@hrydgard commented on GitHub (Sep 29, 2018):
On Mali, those issues are still there underneath even if you use earlier versions with the stripe bug, right? If so they are definitely a separate issue..
@ghost commented on GitHub (Sep 29, 2018):
PPSSPP 1.6.3 Playstore
Yeah separate issue..
@ghost commented on GitHub (Oct 7, 2018):
Upon testing in my brother phone Kata Selfie Mali-T720 GPU MT6753 Hardware
But in my Cloudfone Excite Prime Mali-450 MP GPU MT6592 Hardware it's fixed the only issue I have is the small black square that also exist in MH3RD HD :/
@unknownbrackets commented on GitHub (Oct 7, 2018):
I suppose that means Mali-T7xx doesn't support what we're trying to do. The weird thing is the lines are pretty much the same, but maybe it's because they're all nan.
-[Unknown]
@marosis commented on GitHub (Oct 11, 2018):
I can confirm it. I have same mali-T720
@MJH1000 commented on GitHub (Oct 13, 2018):
The game can be made playable using this cheat, 'Graphical glitches partial FIX'.
ULES00615.ini
@marosis commented on GitHub (Oct 13, 2018):
Please do from it for me File. I did TXT, paste code, saved, renamed to ini and it is not working
@marosis commented on GitHub (Oct 18, 2018):
All cheats are working from PPSSPP forum but that your not.
@MJH1000 commented on GitHub (Oct 18, 2018):
The cheat does not remove all the graphics glitches, only the worst ones.
@marosis commented on GitHub (Oct 19, 2018):
No. I mean i can't launch it. He is not showing in the emulator
@benderscruffy01 commented on GitHub (Mar 24, 2019):
can the label android be add as this doesnt happen on windows
@LunaMoo commented on GitHub (Mar 24, 2019):
No, it will still happen on windows, the feature fixing this issue is disabled on old gpu's. not just mobiles.
@benderscruffy01 commented on GitHub (Mar 24, 2019):
ok thanks for that
@unknownbrackets commented on GitHub (Mar 24, 2019):
Maybe it makes more sense to close this and create a new issue (maybe referencing a list of the affected games for searchability) about guard band not working on older devices.
I don't have an affected device currently (except Vivante which can't handle a lot...) but some ideas collected from above:
> 0(won't be perfect, but might be better... that said, speed penalty.)-[Unknown]
@ghost commented on GitHub (Aug 9, 2019):
@hrydgard Thanks finally its totally fixed ^^
•Mali-T720 MT8163 Android Phone Tablet