Old Regen pages

AamirM · Post by **AamirM** » Thu Aug 21, 2008 7:23 am

Hi,

These are the debuggers I am working on (read: trying to port in Regen from Gens). But it will take some time. So be patient. I'll make a debug only version of Regen because inserting debug hooks into code makes it very slower and its not suitable for the playing games. But all that after the 0.93 release

.

Not as cool as Nemesis' but I think its what most people need.

Nemesis wrote:....and now I feel like I'm derailing Aamir's thread. Sorry. When someone mentiones debugging, I just can't help but talk about it. I still get a stupid ass grin on my face every time I use a breakpoint.

No problem at all. Its good to know that more people are working on debuggers now. I actually use your debugger sometimes

. Its really cool.

stay safe,

AamirM

Eke · Post by **Eke** » Thu Aug 21, 2008 8:18 am

deleting the config file does not solve the issue, rebooting the computer did it

I hate to do this but need to point some issue with your VDP timings (DMA/FIFO/HBLANK) again:

.Outrunners: corrupted graphics
.Out of this World: stuttering line at the bottom of the password screen
.Doucle Clutch: flickering text on intro screen
.Kawasaki Superbike Challenge (same should happen with Formula One as they use a similar program): graphic garbage at the bottom of the in-game screen , works fine with PAL timings though

most of these are games that break as soon as you fix another one until you find the perfect timing scheme, so good luck

Also, we share a common issue with Micro Machines 2 (I thought I fixed it but didn't): on the level selection screen (once you have picked a character), the image you get is wrong (you get a BIG Micro Machines logo instead of the usual vehicule picture ).
Strangely, this does not happen if you return 0 on all EEPROM reads (which this game use) or if you don't emulate the J-CART so I suspect it has something to do with code timing, looks like VRAM is not correcly updated for some reasons... It's a pain to debug so, if you find the solution, let me know

AamirM · Post by **AamirM** » Thu Aug 21, 2008 9:55 am

Hi,

deleting the config file does not solve the issue, rebooting the computer did it

That is strange. Does it behaves like that everytime?

I hate to do this but need to point some issue with your VDP timings (DMA/FIFO/HBLANK) again:

Acutally, I appreciate it. Don't hate to point bugs

.

most of these are games that break as soon as you fix another one until you find the perfect timing scheme, so good luck

I think its the FIFO and DMA timings that is breaking most of these games. I'll work on them again. But a bit later now.

Also, we share a common issue with Micro Machines 2 (I thought I fixed it but didn't): on the level selection screen (once you have picked a character), the image you get is wrong (you get a BIG Micro Machines logo instead of the usual vehicule picture ).
Strangely, this does not happen if you return 0 on all EEPROM reads (which this game use) or if you don't emulate the J-CART so I suspect it has something to do with code timing, looks like VRAM is not correcly updated for some reasons... It's a pain to debug so, if you find the solution, let me know

That game ran correctly uptil version 0.77 and it had both EEPROM and J-Cart emulation. I remember I made some changes to EEPROM code in 0.85 after which this game got broken. Hopefully I will be able to backtrace the problem. I will look at this problem soon and get back with the results. Also the problem gets fixed if you don't emulate EEPROM at all. Thats the sad part in emulation

. A game works if ten things are broken and doesn't when just one thing is broken.

stay safe,

AamirM

Eke · Post by **Eke** » Thu Aug 21, 2008 10:12 am

Yes, as I said, it works when EEPROM emulation is not correct (i.e read at address 380001 return a fixed or unespected value, 0 or 1 does not matter)

If you trace both code execution, you will see that when valid data is read from EEPROM, this involves additional code execution and will change the overall timing (which is normal). Same thing happen when read from 38fffe (jcart adapter) are not correctly supported.

What is *not* normal is that under this timing condition, VRAM got somehow corrupted or not properly updated

more over, the corruption seems to happen before the intro menu screen, I've done savestates with 2 different binaries (one that emulates eeprom, the other which doesn't) exactly at the moment when the main menu appears and it appeared that after loading the "working version" savestate, the "non-working" version will display the good picture later in the race selection menu.

That is strange. Does it behaves like that everytime?

nope, probably an issue on my PC , don't bother with that

AamirM · Post by **AamirM** » Thu Aug 21, 2008 10:50 am

Hi,

Isn't this the game that does not initializes EEPROM and expects some known values (0xFF) to be already present? (or was it some other?)

EDIT:
It also tries to write to ROM (PC = 0x1E526 onwards).

stay safe,

AamirM

AamirM · Post by **AamirM** » Fri Aug 22, 2008 5:12 pm

Hi,

TmEE co.(TM) wrote:PCM is correct now

PSG seems a bit off compared to FM, higher pitch (don't know how much).

You were right. I made a mistake. PSG is slighlty louder than it should be. Thanks, I will correct it.

stay safe,

AamirM

Nemesis · Post by **Nemesis** » Sat Aug 23, 2008 4:11 am

I've got both the J-Cart and non J-Cart versions of Micro Machines 2. Would it help if I dump the EEPROM from both carts?

AamirM · Post by **AamirM** » Sat Aug 23, 2008 8:47 am

Hi,

I've got both the J-Cart and non J-Cart versions of Micro Machines 2. Would it help if I dump the EEPROM from both carts?

Yes, that can be of some help as this game also requires EEPROM to be initialized with certain values (0xFF) otherwise the player names get corrupted. I also now remember that DinoDini Soccer also wanted SRAM to be initialized with 0xFF to work properly and I never understood why it was the case. Shouldn't that data be random??

stay safe,

AamirM

ElBarto · Post by **ElBarto** » Sat Aug 23, 2008 12:10 pm

Nemesis wrote:I've got both the J-Cart and non J-Cart versions of Micro Machines 2. Would it help if I dump the EEPROM from both carts?

I've dumped them, they are the same.

Christuserloeser · Post by **Christuserloeser** » Sat Aug 23, 2008 12:15 pm

AamirM wrote:You were right. I made a mistake. PSG is slighlty louder than it should be. Thanks, I will correct it.

Hm, - what about an option to set the volume for each channel ? (as I like my PSG loud)

TmEE co.(TM) · Post by **TmEE co.(TM)** » Sat Aug 23, 2008 6:50 pm

Memories have FF in their cells when they're new, or when they're erased.

Captain Boomerang · Post by **Captain Boomerang** » Mon Aug 25, 2008 3:32 am

Hi,

I have a cheap USB gamepad that is semi-Playstation style that has both a digital directional pad and a analog directional pad and I was wondering why I can't use both of them at the same time without needing to reasign the input each time. Other emulators like bsnes have the ability to let me use both directional pads at the same time.

One last thing, I've downloaded the latest beta and I found a couple typos in the history file and I've corrected them for you. You can find the updated history file here.

Keep on truckin' AamirM and keep up the good work!

-Adam

AamirM · Post by **AamirM** » Wed Aug 27, 2008 1:34 pm

Hi,

I will see why is it creating that problem on your controller.

stay safe,

AamirM

Captain Boomerang · Post by **Captain Boomerang** » Fri Aug 29, 2008 9:07 pm

Hi,

Thanks for looking into the controller issue with not being able to use both directional pads at the same time. I've been reading about sound issues here and on the zsnes forums and I'd like to see 44100Hz added for SuperHQ for the sole purpose of testing it exactly against Kega Fusion since it does not support 48000Mz. I'm guessing sound issues between the two would become more audible if both were at the same sound output while in high quality mode. Just an idea but I think it would do a lot of good since I can hear louder differences in the output between Kega and Regen which can lead to others possibly hearing things and falsely reporting them.

Not pushing you or anything like I said it's just an idea to even things up for testing purposes against Kega. Thanks a whole lot man!

-Adam

AamirM · Post by **AamirM** » Sat Aug 30, 2008 7:46 am

Hi,

Hi,

Thanks for looking into the controller issue with not being able to use both directional pads at the same time. I've been reading about sound issues here and on the zsnes forums and I'd like to see 44100Hz added for SuperHQ for the sole purpose of testing it exactly against Kega Fusion since it does not support 48000Mz. I'm guessing sound issues between the two would become more audible if both were at the same sound output while in high quality mode. Just an idea but I think it would do a lot of good since I can hear louder differences in the output between Kega and Regen which can lead to others possibly hearing things and falsely reporting them.

Not pushing you or anything like I said it's just an idea to even things up for testing purposes against Kega. Thanks a whole lot man!

-Adam

Looks like a nice idea. I think I'll eventually do it. But there are lots of other higher priority things than this right now so don't expect it anytime soon.

stay safe,

AamirM