PJ 64 and geforce 66gt error

[ryosableng]

Hello there.
I have just upgraded my graphic card from mx4000 to gf 6600 gt.
My spec is barton 3000+ with 1gb ddr2700 ram, win xp sp2.
And i use forceware 91.31 the newest driver that available.
I use PJ64 1.6 with the newest rdb.
For plugin i use Jabo direct3d8 1.6 with this setting:
windowed resolution : 640x480
full screen resolution : 1024x768 16-bit 60hz
full screen sync : tripple buffer Vsync
anisotropic filtering : 16x
full-scene antialiasing : 4x
brightness : 100%
super 2xsal textures : on
always use texture filter : off

The problem is when i switch the screen from window to full screen using alt+enter. When i switch the screen for the first time to full screen there is no prob it went smoothly. But when i switch back to window using alt+enter from full screen then suddenly my screen went messy even i can't see anything on my screen. This happen too if i close using alt+f4 when in full screen.
The screen is clutter so i have to hard restart it bcos i can't see anything at all. I don't have this prob with my old mx4000. And i have tried this on my 2nd rig that use radeon 9200se there isn't any problem like this happen.
And this problem occur only if i use the full-scene antialiasing : 4x
When i turn off the option the glitch wont happen. It can switch normally.
Can somebody help tell me what's wrong with this? And the way to fix it?
I can't figure it. Need help.
Thanks. :flowers:

 

[PsyMan]

Apparently it's a driver issue. You can try using newer or older drivers for your video card. If you don't want to do it (or if this does not fix the problem) then just turn antialiasing off and use a higher resolution.

 

[squall_leonhart]

Apparently it's a driver issue. You can try using newer or older drivers for your video card. If you don't want to do it (or if this does not fix the problem) then just turn antialiasing off and use a higher resolution.

actually its a plugin issue.

i've had the same thing on my FX card for ages since the original drivers for this card.

the way to fix it is to not load a savestate in fullscreen, do it in a window then toggle fullscreen with alt+enter.

you can save in fullscreen, just don't load in it.

i've already been through this with smiff Psyman, the drivers are not at fault, its the plugin that doesn't release the api properly. the DX6. and ricevideo plugin both don't have this issue. and neither did the Dx7 plugin. the problem is that the plugin is calling the nvidia driver to unlock the api and set back to window mode, the reason why older video cards don't get this is becoz they don't support the DX8 features.

i've reproduced this across platforms, across video cards,

the reason why radeons aren't affected is becoz the plugin is made on a radeon i believe :|

to the original user
don't alt+f4 directly out of a game if you've loaded a quicksave in it.

note, that you have to get Pj64 to close completely after loading a rom while in full screen and your able to get back to the rombrowser as you will get a device lost error.

yup just confirmed my work around.
IF > Load rom to fullscreen then load savestate, you may have a problem exiting full screen.

to work around this, DON'T alt-f4 out of pj64
or else you will need to reset

instead, when exiting

either
F12 to end emulation - or alt-enter back to window mode.
(it'll still remain as a full screen image as the api isn't being released from full screen mode)
at this point ALT-F4, out of pj64 and it will properly release the api.

i've discussed with a few beta testers this bug, and it appears to be fixed in the 1.7 plugin.

 

[gandalf]

I don´t have that problem with the 6600GT and 93.31 drivers.
Try changing the color deep from 16 to 32-bit (in descktop and emulator)

One recommendation, enable 8xSAA for PJ64, you must do it from Nvidia control panel, trust me, it´s MUCH better than 4xAA and without framedrop

 

[squall_leonhart]

doesn't help for me gandalf.

its strange.. it doesn't do it for some. but does it for others. and it is related to the plugin as the others don't do this.

 

[PsyMan]

You know... 5900XT and 6600GT are 2 entirely different video cards based on entirely different cores. In other words, if an application has problems with one of them does not mean that it will have problems with the other one too.

Needless to say that the comment you made yesterday was pointless and misleading. In other words, if you don't know about something it's better to just shut your mouth up instead of giving false information (as you do most of the time).

Edit: Yes, this is a warning.

 

[squall_leonhart]

i know the problem exists and i know it has nothing to do with drivers, so push off until you can tell me otherwise.

i've stated how to work around this limitation of the plugin and thats all there is to it.

ALSO

the 5900xt and 6600gt both work on the same basic principle and the core difference doesn't change anything.

seeing as i have stated this also happens on a TI 4800SE WHICH btw, is a Dx8 supporting hardware.

oh btw, if it was false information Psyman, it would do this on all plugins, but no. Rices video in DX8 mode is perfectly fine.

all my games EVERY SINGLE ONE OF THEM, plays fine with ALL drivers, and this problem is relegated only to Pj64.

hence. IT IS a problem with the plugin. as other plugins don't get this.
hopefully, though, it seems the new graphics plugin in 1.7 will fix this as i have already had a guy run through this and tell me its ok.

ALSO i have isolated the problem further by making note of the fact that all the crash logs state that the error occurs in the Dx8 plugin's .dll

if it was the driver, it would probably crash the system.

btw

[4/12/2005 12:48:38 AM] Failed to create device code = 88760868
[4/12/2005 12:33:50 PM] Failed to create device code = 88760868
[9/24/2005 12:58:10 PM] Failed to create device code = 88760868
[9/24/2005 12:59:47 PM] Failed to create device code = 88760868
[9/24/2005 1:25:06 PM] Failed to create device code = 88760868
[9/24/2005 2:10:43 PM] Failed to create device code = 88760868
[10/5/2005 4:25:46 PM] Failed to create device code = 88760868
[9/14/2005 3:20:40 PM] Failed to create device code = 88760868
[9/14/2005 4:17:07 PM] Failed to create device code = 88760868
[9/14/2005 4:36:07 PM] Failed to create device code = 88760868
[1/14/2006 10:14:50 AM] Failed to create device code = 8876017C
[1/14/2006 10:14:50 AM] Failed to create device code = 8876017C
[1/21/2006 12:51:30 PM] Failed to create device code = 88760868
[1/25/2006 6:26:01 PM] Failed to create device code = 8876017C
[1/25/2006 6:26:11 PM] Failed to create device code = 8876017C
[1/25/2006 6:27:19 PM] Failed to create device code = 8876017C
[1/25/2006 6:27:22 PM] Failed to create device code = 8876017C
[1/25/2006 6:27:30 PM] Failed to create device code = 8876017C
[1/25/2006 6:27:59 PM] Failed to create device code = 8876017C
[1/25/2006 6:28:03 PM] Failed to create device code = 88760868
[2/25/2006 1:49:57 PM] Failed to create device code = 8876017C
[2/25/2006 1:49:57 PM] Failed to create device code = 8876017C
[2/25/2006 1:52:28 PM] Failed to create device code = 8876017C
[2/25/2006 1:52:28 PM] Failed to create device code = 8876017C
[2/25/2006 1:57:06 PM] Failed to create device code = 8876017C
[2/25/2006 1:57:07 PM] Failed to create device code = 8876017C
[2/25/2006 1:57:25 PM] Failed to create device code = 88760868
[2/25/2006 2:01:59 PM] Failed to create device code = 8876017C
[2/25/2006 2:17:05 PM] Failed to create device code = 88760868
[2/25/2006 2:17:19 PM] Failed to create device code = 88760868
[2/25/2006 3:52:11 PM] Failed to create device code = 88760868
[4/8/2006 8:34:21 PM] Failed to create device code = 88760868
[4/8/2006 8:35:05 PM] Failed to create device code = 88760868
[4/8/2006 8:35:37 PM] Failed to create device code = 8876017C
[4/8/2006 8:35:37 PM] Failed to create device code = 8876017C
[5/20/2006 11:22:02 AM] Failed to create device code = 8876017C
[5/20/2006 11:22:12 AM] Failed to create device code = 8876017C
[5/27/2006 7:47:36 AM] Failed to create device code = 8876017C
[5/27/2006 7:49:10 AM] Failed to create device code = 88760868
[5/27/2006 7:50:20 AM] Failed to create device code = 8876017C
[5/27/2006 7:50:25 AM] Failed to create device code = 8876017C
[5/27/2006 7:50:56 AM] Failed to create device code = 8876017C
[5/27/2006 7:51:04 AM] Failed to create device code = 88760868
[5/27/2006 7:53:49 AM] Failed to create device code = 8876017C
[6/3/2006 12:11:01 AM] Failed to create device code = 8876017C
[6/3/2006 12:11:33 AM] Failed to create device code = 88760868
[6/4/2006 1:51:15 AM] Failed to create device code = 8876017C
[6/4/2006 1:51:44 AM] Failed to create device code = 88760868
[6/18/2006 11:18:18 PM] SetViewport (scissor) failed 8876086C (-2147483584,64,-2147482688,645)
[6/18/2006 11:18:18 PM] SetViewport (scissor) failed 8876086C (-2147483584,64,-2147482688,645)
[6/18/2006 11:18:18 PM] SetViewport (scissor) failed 8876086C (-2147483584,64,-2147482688,645)
[6/18/2006 11:20:12 PM] Failed to create device code = 8876017C
[7/17/2006 4:10:18 AM] Failed to create device code = 8876017C
[7/17/2006 4:10:24 AM] Failed to create device code = 8876017C
[7/17/2006 4:10:34 AM] Failed to create device code = 88760868

is the error i get after reproducing this bug. this is becoz the plugin hasn't released Dx8 properly in the nv drivers.

but closing Pj64 completely fixes it.

here is another error i get that Windows Error reporting mentions

Event Type: Error
Event Source: Application Error
Event Category: None
Event ID: 1000
Date: 25/02/2006
Time: 12:57:21 PM
User: N/A
Computer: ATHLONXP2000
Description:
Faulting application project64.exe, version 0.0.0.0, faulting module jabo_direct3d8.dll, version 1.0.0.1, fault address 0x000017e9.

For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.
Data:
0000: 41 70 70 6c 69 63 61 74 Applicat
0008: 69 6f 6e 20 46 61 69 6c ion Fail
0010: 75 72 65 20 20 70 72 6f ure pro
0018: 6a 65 63 74 36 34 2e 65 ject64.e
0020: 78 65 20 30 2e 30 2e 30 xe 0.0.0
0028: 2e 30 20 69 6e 20 6a 61 .0 in ja
0030: 62 6f 5f 64 69 72 65 63 bo_direc
0038: 74 33 64 38 2e 64 6c 6c t3d8.dll
0040: 20 31 2e 30 2e 30 2e 31 1.0.0.1
0048: 20 61 74 20 6f 66 66 73 at offs
0050: 65 74 20 30 30 30 30 31 et 00001
0058: 37 65 39 0d 0a 7e9..

oh btw, Psyman, before you open your mouth again
i've tested this on a 6600GT myself when i was building my uncle a computer
GUESS WHAT, same damn bug.

4 different video chips (Ti 4800SE, FX5700, FX5900XT, A6600GT)
30 different sets of drivers (from detonator 50 to forceware 91.31.
drivers set to defaults (no aa, no af, no performance tweaks)
plugin set to defaults. (, no aa, no af, no smoothing)
resolution dropped (fullscreen 800x600 + 1024x768)
all memory resident programs closed
disable all startup programs.

All environments have been checked,

the problem IS THE PLUGIN.

i can reproduce this bug, every single time.

wanna try it yourself?

heres the steps to follow to cause this.

set the Dx8 plugin to any setting you like.

1.start a game and toggle fullscreen.
2. quick save your game.
3. reload that quick save while in fullscreen several times.
4. try to alt-enter or F12 out of emulation.


this bug won't occur to everyone for the one simple reason.
NOT EVERONE USES SAVESTATES AS MUCH AS I DO.

it also only occurs on certain roms.

i've been able to reproduce it everytime on Zelda OOT and zelda MQ, supermario64, donkey kong64.

for some reason these 2 don't.
Zelda MM doesn't.
Banjo Tooie doesn't

so it seems to have something to do with closing emulation on those roms.

i've also noticed that cpu time stays at around 70% usage till pj64 closes.

 

[PsyMan]

OK... I can't reproduce neither of the problems... (actually a friend of mine's can't since I don't own a 6600).

You do realize that ryosableng didn't even mention anything about using savestates, right? He also mentions that he's having the problem only when using Antialiasing 4x.
Have you ever thought that he may not even have the bug you're talking about? I never said that the bug you mention does not exist. I simply don't think (by the info that ryosableng gave us) that he has the same bug.

By the way, the warning I gave you is still there and also has to do with your attitude.

 

[gandalf]

okey.
i start the game (cruis´n world, donkey kong 64, super mario 64)
switch to full screen....yes
savestate before game starts.....yes
load that savestate several times......yes
exit rom/emulator from fullscreen pressing f12 or alt+f4 with that bug......no


i thinking, that it´s driver problem, maybe with the chipsets drivers

 

[squall_leonhart]

no, its not.

did you try the games i said do it.

they don't always do it either.

its odd,

btw, psyman, the error is very similar to mine, and mine is only caused by quickloading a state in fullscreen, if i load it in window mode, theres no problem

still i've worked out a work around and my workaround might work for the original poster as well.

he hasn't posted back to inform us.

btw.. sorry for any attitude.. i haven't had a good month, ... bad breakup does that....

 

[ryosableng]

Thank for the advices guys.
I have found 2 way to solve this problem. And why it seems that it only affects me aside from the other members that are using 91.xx driver.
First way is changing the forceware driver from 91.xx to 8x.xx. I have tried 91.33 driver and the result is same with 91.31 that i've tried before. But when i switch to 84.21 the problem is solved. It seems like all the driver that start from 9x.xx and higher series have this problem.
Second way is changing the desktop color from 16bits to 32bits when using 9x.xx forceware and the problem is no longer present. I think that's why the others members that is using 9x.xx don't encounter this problem bcos maybe they're using 32bits desktop color depth. I use 16bits color bcos i'm still playing ragnarok online games, and that made me easier to use alt+tab to switch to desktop bcos that damn game don't have 32bits color option.
I think i've cleared the problem and maybe this will help others that encounter the same problemo.
Thankiu.

 

[ryosableng]

Can u guys confirm about the desktop color depth problem?
I wanna know if the problem happen only on me or not.
Can u guys test it with using 16bit desktop color depth?
Are the result is the same as me or not.
Please post ur testing here so we can confirm it if the problem are real.
Thanks.