Weird Issue with Star Ocean on SD2SNES

Hm, I'm at a loss right now, I'll continue trying to reproduce it with different kinds of SNESes.
I confirm the MD5 sum of the patched ROM.

What are your SNES revisions (CPU/PPU1/PPU2)?

 

I just got it to happen when a save is already present. It took quite a long time. I let it sit at the save screen while watching a movie, it may have taken more than 20 minutes to show up.

CPU:2 PPU1:1 PPU2:3
It's a Super Famicom, if that matters at all.

 

IIRC my SNES (eu model) is 2/1/3 too, I can't check atm as I am installing a superCIC :encouragement: (thanks for that too ikari_01!)

 

I'll try it too when I get home from work to see if it happens on the Japanese version, and if not, I'll try to make a fresh English translated version and try that. I'm presently playing Dragon Quest IV (trying to play the whole series in order, and this is the only one of the first six not released for SFC, so I'm emulating the PSX version), so I'll have plenty of time to just leave my Super Famicom on.

 

Mines is a CPU Rev:2 PPU1 Rev: 1 PPu2 Rev:3.
Its an NTSC/U SNES jr.

 

anyone test it on bsnes

 

The translated rom works quite well on BSNES the problem is that in order to play it on SD2SNES you need to apply yet another patch that decompresses the game rom into a 12MB rom. I think that is where the problem lies.

 

If that is true then the problem should occur on the PowerPAK and Game Doctor SF7 too.

 

^ I am currently testing on the PowerPak, I am just leaving it at the save screen and see what happens

edit1: 25mins of waiting, nothing happens , still waiting...

edit2: 1hr 10mins later and still no error... going to reset now and do another wait cycle ... btw my pal snes is 2/1/3

edit3: another 15mins and no error, gonna stop for today, need to play something LOL

 

I went ahead and went straight for an English translated version. I patched the Star Ocean rom from the No-Intro set (which, going by the CRC32, seems to be identical to "Star Ocean (J) [!].smc" from the GoodSet) with soe.ips obtained from zophar.net, then with 96Mbit_SO_ENG.xdelta from romhacking.net. I'm more than 40 minutes in on that save screen with no saves, and I want to say that I don't know what you guys did, but try doing it differently.

I'm not sure why you put a header on, for one. I couldn't find any instructions for either patch that said to use a headered rom (correct me if I'm wrong), and it's not 1998 anymore, so why would you need a header, unless the patch was wonky enough to have been made from a headered rom? Anyway, yeah, try not doing that, then report on whether you still see the same problem.

I will try the Japanese version now and report.

edit: Erm, CRC32 is the same. Still, I don't see why you would add a header...

edit #2: Trying the Japanese version now. About 10 minutes on the save screen and still nothing unusual.

edit #3: Still nothing more than 30 minutes in. Aborting. Could it be a hardware issue on certain SD2SNESes then? Seems to be affecting an awful lot of people for that to be the case though.

 

I added a header because after I patched it headerless using the 96 Mbit patch I got from romhacking.net it wouldn't run on my sd2snes at all. Just a black screen. There were to instructions pertaining to header or no header. All I know is when I did it with no header, it didn't run at all.

The first time I tried it, I know I let it sit for more than half an hour, and nothing. But I have also gotten it to glitch after around 5 and 20 minutes two separate times. Whether a save is present or not doesn't seem to matter.

EDIT: It's possible I might not have actually added a header and I could be thinking of something else. I can't remember now. But how much of a difference would that make anyway? I know if patches that require headers are applied to headerless Roms, they typically don't work at all. Why would this one thing be wrong but everything else seems to work fine? I will try repatching everything when I get home (working) and report my findings.

 

Yeah, I don't think the ips patch would work with a headered version anyway. I wasn't sure about other patch formats though, and whether there might be something that auto-detects it. In any case, definitely unnecessary.

How reliably can you guys get this to work? I decided to try for a fourth time.

 

Other than the first time I tried, it glitched every time. How long it took to glitch seemed like the only variable to me. The second time took less than a minute. Third took around 5 and the fourth took around 20. So its possible, if I waited long enough, that the first time I tried would have also glitched.

What is the crc of your patched ROM? Does it match the crc of the one that we've been having problems with?

 

6F3EAC5A. Same as you guys'. I ran it from the root, in case that makes any difference. It's now been almost 30 minutes, and still nothin'. This time, what the hey, I'll leave it on all night to see what happens, if I have to.

 

I Just did exactly as you described and I still get the issue. Maybe my SD2SNES is defective or something...

 

yeah I never added headers or anything, I patched like Eyedunno.

BUT for the powerpak to understand the game is a sram-saver one has to edit 4 bytes (it runs without editing it just does not save properly). The same 4-byte-patched rom does not work on the sd2snes , it just boots to a blank screen.
Maybe the save screen thing is related to it, some weird bug or something... I know I said it before and this is just speculation but maybe it went unnoticed the 1st time. Why sd2snes does not run the powerpak version that tells that the game is a sram saver?

 

The patch is just there to make the header information comply with PowerPAK's memory map / cartridge type detection, no code patches involved. sd2snes detects it in a different way so it will try to map the PowerPAK patched one as a regular HiROM which fails obviously. That aside it's completely unrelated to the issue at hand.

I still can't reproduce it after 10 runs of about an hour each. ;_; However I don't have a KRIKzz made sd2snes to test it with, only a self made one. His sd2sneses turned out to be excellent tools in pointing out timing flaws in the design I don't know why, maybe a different batch of FPGAs.
I might be on to something though: After loading the SRM file but before actually running the ROM I took a 256 byte chunk of SRAM as a sample for testing purposes and constantly compared it with the current SRAM contents while the game was running. Live SRAM content should never differ from the sample as long as the game isn't saved. But when entering the title screen or the save menu there's a number of mismatches, just for a millisecond or so.
Could be a red herring, I don't know if Star Ocean writes stuff to SRAM temporarily. I'll try and debug that later today.

 

ah, thanks for explaining the powerpack patch ikari_01 ...
I have a KRIKzz sd2snes btw ....

 

I also have a KRIKzz one (one of the earlier batches, I'm sure - I was in Stone Age Gamer's first shipment). I did leave it on at the save screen all night. No problems to speak of. Could a later batch have been slightly different, perhaps in terms of how it handles SRAM?

 

I'm pretty sure all of the deluxe edition pre-orders were made using PCBs from the same batch. They were broken up into waves to accomodate production and assembly of the carts and other stuff involved with the "deluxe edition." Maybe KRIKzz or StoneAgeGamer could confirm this.