ata4 / glrage Goto Github PK

When you end the level in Tomb Raider and Tomb Raider: Unfinished Business while holding the Action button, the game freezes and the only thing you can do with it is to kill its process.

Binary diff for a tombati.exe patch that fixes this:

0001CC27 0D 33
0001CC28 E8 90
0001CC29 24 90
0001CC2A 0D 90
0001CC2B 00 90
0001CC2C 00 90
0001CC2D 85 90
0001CC2E 35 90
0001CC2F F4 90
0001CC30 EE 90
0001CC31 45 90
0001CC32 00 90
0001CC33 75 90
0001CC34 F3 90
0001CC5A 2F CD

…and here's a tombub.exe patch:

0001C987 0D 33
0001C988 E8 90
0001C989 74 90
0001C98A 0C 90
0001C98B 00 90
0001C98C 00 90
0001C98D 85 90
0001C98E 35 90
0001C98F B4 90
0001C990 E7 90
0001C991 45 90
0001C992 00 90
0001C993 75 90
0001C994 F3 90
0001C9BA 2F CD

Save files you can test the patches with:

https://0x0.st/z-WM.1
https://0x0.st/z-Wu.1

Hi. Do you still have the original ATI3DCIF to OpenGL wrapper? It was uploaded to https://sourceforge.net/projects/glcif/ but the site is down and I can't reach the original poster/coder "msulak". Thanx!

Here's my preliminary take on rendering issues:

Files between PSX and ATI versions are binary identical for the ships ALLSH.PRM/CMP, it's quite puzzling as on why some of the ships are picking incorrect textures though they are correctly drawn, I guess that's unimportant for our problem.

SJR format are header-less RAW RGB555 textures, their width is either 256 or 512 pixels, height is probably inferred using width and file size, their width is hard-coded in the game for those I've seen.

Now for the missing polygons,

1. Their visibility seems tightly related to the refresh rate the game is currently at, going to windowed mode changes somewhat what's being drawn.
2. If you closely look at the starting lights in Altima VII, they are correctly rendered in their initial state (gray color). But as soon as they get tinted (red, orange, green) they do disappear. (shader issue ?)
3. There's a lot of jitter in regards to when a polygon will be drawn or not, stopping at the billboard before the long slope in Altima VII and looking at the right part of the billboards (made of 2 quads) will intermittently draw it. This game uses integers for its data and PSX did too (a rough guess) for 3D calculation, could there be some unintended Z-fighting / culled polygon being introduced while bringing the data from the game and transforming it for the OpenGL (being way more picky than a PSX) pipeline ? (could some fixed point arithmetic in the shader be of any help ?)
4. Artifacts definitely change when slowing down the whole thing (a poor man approach using Sleep in ContextImpl::renderBegin()) and you effectively see correct rendering (mostly). I am starting to wonder that this game might need a slow down (missing ATI logo, samples playing at wrong times) fix but I reckon this a tough/time consuming job.

Or a combination of multiple factors occurring at the same time ?

An interesting challenge (I guess) would be to try to get those quads presenting the text (in wireframe mode) to be really straight, as of now their drawing is missing many parts.

What about the other CIF games, do they exhibit the same sort of rendering issues ?

Sorry if it's a bit scattered through, I'll continue investigating and see if I can make any progress on the topic ... maybe that'll give you an idea or two and you might come up with a smarter approach 🎯

i5 4690k
GTX 970
Triple Monitor /w nVidia Surround @ 5860x1080(BC)

I admit that I have not tested any other mods, or even "vanilla" OpenGlide yet.

I'm experiencing some really weird frame-rate issues.
Game is fine for awhile, then it starts hitching. Seems like maybe it's related to large areas, or may something to do with how much of the game gets loaded in as I'm playing. Like, maybe it's a bug from being an older game. I know it's capped at 30 FPS. That would be fine if it would deliver even frame-pacing. The weird part is how I'll run past part of a level the first time, and it's fine, but when I come back and face a certain direction, it stutters! Next time I test it, it'll do it in a different spot! Sometimes I can kind of 'kick' the game by enabling/disabling a frame-rate limiter, or swapping between window/fullscreen, but the hitching ALWAYS returns. It's so weird?? I've theorized that maybe parts of the level get...corrupted upon being loaded? Like, it's unoptimized (the game, not your mod)? Then if that part of the level gets cleared and reloaded, then problem comes back in a different spot? I'm just making stuff up by this point to try and understand the intermittent nature of these symptoms. I love this game, and I love these mods, but if I can't get even frame-pacing with 30 FPS on my 60 Hz monitors, then I give up.

I've tried using every single combination of the following, sometimes simultaneously:
RTSS
NVCP
SpecialK
RivaTuner
(Above utilities used to either limit framerates or force VSync or 1/2VSync On/Off)
In-Game VSync on/off

Another suspicion I have is that the application window that it runs in is somehow being isolated off in its own little world away from WDM, or however this stupid composite desktop is being rendered in newer versions of Windows. I've tried with/without 'Full Screen Enhancements' but it had no effect on performance.

I have Visual Studio 2008 and 2010 and 2019.
It opens in 2019 and has tons of bugs (errors)

I'm sorry that this is not a patch issue but a miss feature. I did everything as descrived in your readme, but I would like to maintain my savefiles. I tried to copy and paste them I can see them into the game but when I try to load it, it crashes. Is there a way to maintain saved files?

EDIT: I found a workaround solution. There is this page with saved games in format of this patch: https://raidingtheglobe.com/downloads/tomb-raider-1/60-tomb-raider-1-ati-glrage-savegames

Hi.

When I saw this project, I instantly grab my old CD of Assault Rigs and installed the game. Everything went fine but when start ARIGS.EXE, I got the following error message:

My system is a Windows 10 PC with 8GB RAM an an Geforce 680 GTX. I tried it on an PC of a friend but the same error occurs.

Is it a problem with the wrapper or are some of the Files on the CD maybe broken?

Hello,
I launch GLRage on Linux with Wine 2.x.
It works very good on Wine with some little tweaks.
Last time I replaced all RPL cinematics with AVI in HD and GLRage does not play them and send me errors messages:

So I resume: on Linux with WIne, Glrage play RPL videos but not .avi or other videos formats.
What does it uses to play/decode video?
Do I have to install a special codec or dll on WIne?

Thank you for your help.

This is potentially a bug in wine or in mesa, but I figured I would open up a bug report here first just in case. I'm using r10 here, so it shouldn't be a problem with compat profiles.

I've started a branch for Star Wars: Shadows of the Empire @ https://github.com/aybe/glrage/commits/shadows_of_the_empire and while it doesn't yet work, it does start ... but the list of devices/resolutions is empty.

Do you know where I could find examples I could base myself on for this DirectX 3 game ?

You can get the demo here if you want to give it a try:
https://archive.org/details/SHADOWS_201406

EDIT: I use the demo for now because it's simpler than the original which has 16-bit setup and CD protection ... a bit cumbersome for now.

The game crashes right after playing the video:

(actually it crashes 2 times in a row and pressing the Debug button does not start Visual Studio)

Faulting application name: WIPEOUT.EXE, version: 1.0.0.26, time stamp: 0x32134bd7
Faulting module name: WIPEOUT.EXE, version: 1.0.0.26, time stamp: 0x32134bd7
Exception code: 0xc0000005
Fault offset: 0x0006a09f
Faulting process id: 0x2dac
Faulting application start time: 0x01d15473af15d303
Faulting application path: C:\3D Rage Games\WipEout\WIPEOUT.EXE
Faulting module path: C:\3D Rage Games\WipEout\WIPEOUT.EXE
Report Id: 0609fa33-642d-4325-bd7f-a41a5e458309
Faulting package full name: 
Faulting package-relative application ID: 

Faulting application name: WIPEOUT.EXE, version: 1.0.0.26, time stamp: 0x32134bd7
Faulting module name: WIPEOUT.EXE, version: 1.0.0.26, time stamp: 0x32134bd7
Exception code: 0xc000041d
Fault offset: 0x0006a09f
Faulting process id: 0x2dac
Faulting application start time: 0x01d15473af15d303
Faulting application path: C:\3D Rage Games\WipEout\WIPEOUT.EXE
Faulting module path: C:\3D Rage Games\WipEout\WIPEOUT.EXE
Report Id: 264bb75a-df8f-4c7d-818f-365450865365
Faulting package full name: 
Faulting package-relative application ID: 

My configuration:

Windows 10
Core i7 4770
16 Gb RAM
ATI HD7850 2GB

Let me know if there's anything I can do over here to provide more details.

Thank you!

I found some issues with looping ambient audio tracks when music_mute_menu is enabled in tombati. When you leave the game paused long enough, you'll notice that the next iteration of the track is unmuted. When that iteration finishes, the track stops for good, even when you unpause the game. I traced the problem to this code in musicPlay:

    // this one will always be set back to true on the next tick
    *m_tombCDLoop = FALSE;

    // Calculate volume from current volume setting. The original code used
    // the hardcoded value 0x400400, which equals a volume of 25%.
    uint32_t volume = *m_tombCDVolume * 0xffff / 10;
    volume |= volume << 16;
    auxSetVolume(*m_tombAuxDeviceID, volume);

When the game is paused, m_tombCDLoop is never set back to true, which is why the track eventually stops. Setting the volume unmutes the track.

When I take this code out the problem is fixed, and no new issues seem to arise, i.e. no tracks loop that aren't supposed to, and the volume setting is still maintained. I'm using the WinMM wrapper with FLAC audio tracks you posted on tombraiderforums.

Greetings from VOGONS crew :) This isn't really an issue per se.

Your project was mentioned in this thread: http://www.vogons.org/viewtopic.php?f=9&t=36276

In case you have never seen it, Gona left this wonderful repository for ATi 3D Rage and Rage Pro fans:
http://gona.mactar.hu/ATI_3D_CIF/
It lists compatibility of different games with different cards with different drivers, plus links the ATI 3D Rage Pro SDK. :)

Hi. First of all thanks for your wonderful wrapper. I have one request for you. Is it possible to add support for PC version of Resident Evil 2, which is also support for ATI 3D Rage Pro series and some 3D Rage LT according to game readme. But when I try to launch this game with your wrapper I have error message "Pointer to original DirectDrawEnumerateA function not found". I hope there's a way to make RE2 works with your wrapper, thank you.