PS2 [Open PS2 Loader] Game Bug Reports

You can try separate modes e.g. 1, 2, 3...6, 8. Mode 7 is unused.
You can try combinations 1+2+3+6, 1+3+6, 2+3+4+6...
Maybe eventually something might help, but it might be also a problem with initialization check.
Ok, I tried all mode to be on and now it seems like game is running.
BUt came across another issue, mic is detected, but when I go into tutorial to speak to my mic, no signal is being detected by game
 
Can you at least try to turn off mode 4+7?

Can you try 2+3+6, 1+3+6 or only 2?
@homebro2000 Please, try these combinations and tell us which one works:

- Mode 1+6
- Mode 6+8;
- Mode 6+3+8;
- Mode 1+6+8;
- Mode 1+3+6+8;
- Mode 2+6+8.

I think that soon he will have logo of this game remembered for ages,
when he will boot this game for a 100 time.
8d2ad9671de3.jpg
 
I think that soon he will have logo of this game remembered for ages,
when he will boot this game for a 100 time.
8d2ad9671de3.jpg
We need to get to the bottom of this. Besides, only 6 more tries to check those combinations. If he's so deep already, it's worth it.
 
@Grahf
poles apart,if your "accurate reads for CD" finally only for one or two games,i think make patch for every game is a best solution
however,please show your implementation not have empty talk
 
What do you mean with "a patch for every game"? Nobody is going to debug Shadow Man 2 and the PS remake. It didn't happen so far for a reason.

We had special modes just for single games in the past and now it's suddenly bad to add alternative speeds to Mode 1 in order to fix two games?

I'm all out for it. I'm not an OPL developer, so I can't add it myself, but those who can: please add three speed options for Mode 1 (850 kB/s, 1200 kB/s and the current speed for DVD games).

Regarding Racer Revenge, who said this game is broken to begin with!? The music cuts in and out like that even from the retail disc. It seems this game is designed this way. Even the PS4 re-edition behaves like that.
 
Last edited:
Shadow Man 2econd Coming seems to have all I/O done by its sound driver, even for reading game files.

What do you mean with "a patch for every game"? Nobody is going to debug Shadow Man 2 and the PS remake. It didn't happen so far for a reason.

Magic might happen, if you asked for it. But I'll do it, since people have been asking.

I doubt I can fix the PS remake because it is a runtime bug, something which seems to happen during rendering, likely as an indirect effect of reading. I have very little experience with debugging rendering and sound-playback issues.
I might need to understand how the game works, dump VRAM to see where it goes wrong and then create a patch. But that's too much work, when MIPS (and VU) assembly is the only thing that can represent the whole game.

If only 2 games can be fixed by lower the transfer rate by that much, then I would rather write game-specific fixes for them because the setting would be useless for nearly every other game.
Even if we cannot debug and solve the actual underlying problem, then an additional delay/additional configuration of mode 1's speed limit can be added like what Jimmikaelkael did for Kingdom Hearts 2. The user would hence have less settings to change, and it's possible to blacklist because we already know that these games have problems.

Worth mentioning that the CD-ROM speed in PS2 as it is widely-documented only applies to reading data from the centre continuously.

It has documentation for reading performance from both inner-most and outer-most tracks, but none of the rated speeds are that low.

However, due to seek times which can be very high if a game uses a ton small files (like SD textures), the timing becomes very tricky. Since we can't emulate the actual movement and data reading of a laser assembly,

Games have to be designed to reduce seeking, as seeking is documented very slow and so is the IOP, so I don't think it is really a problem. But if you can come up with a model for this and it actually helps, it can be done. Software is flexible.

BTW Castle Shikigami 2 needs Mode 1 to have BGM even for USB, but I don't know why?!
I mean normal Ifcaro OPL with transfer speed "set for DVD".
No other modes will helps this game to have BGM?!

It's because this game's sound driver monitors sceCdReadPos() to check for new progress in reading the audio stream.
So if it is either read too quickly or not updated incrementally, the sound driver fails to continue playback.

Other than limiting speed, accurate reads (mode 1) also:
  • Imposes a 5000-us delay between the completion of a command and the interrupt callback. This is for compatibility for some games that were designed to assume that the callback does not occur immediately after the command is issued.
  • Limits the transfer block size to 8 sectors. While I have no information on the CD/DVD's host controller capabilities, I think I came up with this number from the Sony CDVDFSV's buffer size.
This also solves performance problems in some games (i.e. Dragon Quest 8), due to them requesting a lot of data, which the HDD unit can satisfy quite quickly... causing the active thread to not sleep for extended periods of time. This mode solves that problem because it breaks up the long data transfers, allowing lower-priority threads to get CPU time.

This compatibility mode was originally intended to replace the user-setting of DMA modes, but it somehow did not help all games and I do not know why, so the DMA setting remained (indefinitely) as a stop-gap measure. The DMA modes affect the transfer rate between the PS2 and the HDD, but it was for device compatibility and not as a measure for software compatibility.

I turned on everything 1 to 8
Do not do this. Some modes were meant for specific types of games or specific uses.
For example, DVD-DL emulation was meant for flattened images. It will cause games on real DVD-DL images to malfunction.
 
Last edited:
Magic might happen, if you asked for it. But I'll do it, since people have been asking.

I doubt I can fix the PS remake because it is a runtime bug, something which seems to happen during rendering, likely as an indirect effect of reading. I have very little experience with debugging rendering and sound-playback issues.
I might need to understand how the game works, dump VRAM to see where it goes wrong and then create a patch. But that's too much work, when MIPS (and VU) assembly is the only thing that can represent the whole game.

If only 2 games can be fixed by lower the transfer rate by that much, then I would rather write game-specific fixes for them because the setting would be useless for nearly every other game.
Even if we cannot debug and solve the actual underlying problem, then an additional delay/additional configuration of mode 1's speed limit can be added like what Jimmikaelkael did for Kingdom Hearts 2. The user would hence have less settings to change, and it's possible to blacklist because we already know that these games have problems.
Just checking if I understand this correctly.
Are you saying that you will try to debug PS gen1 to fix the actual issue with it, even though you believe it to be too much work.
If you are unable to fix the game you plan to make a game-specific lower transfer rate?

If that is correct, is there any way I can help you debug it and would you like me to find the highest speed that it actually works with?
 
Shadow Man 2econd Coming seems to have all I/O done by its sound driver, even for reading game files.

Magic might happen, if you asked for it. But I'll do it, since people have been asking.
Why would you waste your time, especially since you said it yourself that you have very little time to do this, in order to patch two games which already work with these lower speeds? I don't know if you realize this, but I've compared the load times for these games between retail discs and OPL with 850 kB/s and 1200 kB/s - they are about the same, so we don't lose performance in any visible way. These speeds are perfect and it's just semantics whether you're going to call them a patch or a workaround. Besides, I assume they smooth out many other CDs game, preventing random freezes and/or glitches, but that needs extensive gameplay tests for a ton of games I don't even own. The good thing about this, though, is that there's no risk involved in introducing two lower speeds.
I doubt I can fix the PS remake because it is a runtime bug, something which seems to happen during rendering, likely as an indirect effect of reading. I have very little experience with debugging rendering and sound-playback issues.
I might need to understand how the game works, dump VRAM to see where it goes wrong and then create a patch. But that's too much work, when MIPS (and VU) assembly is the only thing that can represent the whole game.
It's worth adding just for one game, in all honesty. There's next to no development work involved since it's just a value change for Mode 1. The users already have relatively complex features built-in, like GSM. Why would adding three speed values to Mode 1 be a problem when the settings menu is already cluttered with other stuff and nobody seems to mind that?
It has documentation for reading performance from both inner-most and outer-most tracks, but none of the rated speeds are that low.
Theoretical reading speeds for continuous files.
Games have to be designed to reduce seeking, as seeking is documented very slow and so is the IOP, so I don't think it is really a problem. But if you can come up with a model for this and it actually helps, it can be done. Software is flexible
Multiple PS2 developers said that seeks times are a killer on that system. Optimization is not only possible, but simply needed. There were many attempts to reduce the load time as much as possible, but it was always a problem because you are dealing with a laser assembly which cannot seek as fast as an HDD.
The DMA modes affect the transfer rate between the PS2 and the HDD, but it was for device compatibility and not as a measure for software compatibility.
The DMA are needed for both. Nobody can account for all possible HDD models out there and some games simply benefit from certain DMA modes. It would be absurd to removes such a great and flexible feature.

Also, Racer Revenge needs to be removed from the list of broken games. The BGM cutting in and out so abruptly is an integral part of that game. I happens identically from a retail disc and even in the PS4 re-release.
 
Last edited:
The 24x CAV mode has got the maximum data transfer rate at the outer edge (3,6 MB/s). At the centre of the disc it would be something about 1,5 MB/s. If we assume, that the developers mistakenly put the data close to the inner edge, 850 kB/s is still nearly double slower than a minimum read speed. I don't think it is a close replication of the CDVD behaviour.
 
The 24x CAV mode has got the maximum data transfer rate at the outer edge (3,6 MB/s). At the centre of the disc it would be something about 1,5 MB/s. If we assume, that the developers mistakenly put the data close to the inner edge, 850 kB/s is still nearly double slower than a minimum read speed. I don't think it is a close replication of the CDVD behaviour.
1200 kB/s (this speed is great for Shadow Man 2 via HDD - only SMB requires a lower speed) and 850 kB/s are close enough from HDD and SMB. That's because the seek times on a CD are a bitch and you're not even considering the error correction inherent to optical drives. You're giving values for maximum theoretical continuous speeds. Console video games with optical drives don't work like that, unless they just stream a FMV file.
 
Last edited:
I've a problem with latest OPL r1214.
I mean I cannot save settings on HDD when there is no MC*.
Without any MC"s, USB, settings will not be loaded from HDD.
I'm booting OPL from HDD.

* - When MC is inserted all setting will be saved on MC not on HDD.

Currently only everything is fine when MC is inserted.

Previously (e.g. OPL 1196) when I booted OPL from HDD without any MC, I can either save or load setting from it (HDD).
Maybe currently new script for recognizing from where OPL was booted is not working fine?

Unfortunately previous version didn't recognize from where OPL was booted,
so MC always has a higher priorytety with saving or loading settings
 
Last edited:
  • Like
Reactions: TnA
I've a problem with latest OPL r1214.
I mean I cannot save settings on HDD when there is no MC*.
Without any MC"s, USB, settings will not be loaded from HDD.
I'm booting OPL from HDD.

* - When MC is inserted all setting will be saved on MC not on HDD.

Currently only everything is fine when MC is inserted.

Previously (e.g. OPL 1196) when I booted OPL from HDD without any MC, I can either save or load setting from it (HDD).
Maybe currently new script for recognizing from where OPL was booted is not working fine?

Unfortunately previous version didn't recognize from where OPL was booted,
so MC always has a higher priorytety with saving or loading settings
rev1213 - SP193 - Changed how devices are selected for loading/storing config files. When loading: 1. Check memory cards. 2. If config could be loaded, try the device that OPL was booted from (supported devices only). 3. If config could not be loaded, try all supported devices. 4. Default to memory card, if no config file could be loaded. - Mon Dec 17
 
  • Like
Reactions: TnA
Back
Top