Genesis Core

[Sorgelig]

there is still many things to fix before think about fine details.

[hyperterminal]

I just played a few more rounds of OutRunners and noticed a regression. When you are taking a fork to the next stage there are vertical bars on the horizon, ais if the ram was faulty. This did not happen on older versions of the core.

[cacophony]

I assume cart saves are not implemented yet? I tried with Sonic 3 & Knuckles and couldn't get it to work.

[the1ross]

Ive been having a blast with this core. Just wanted to report some bugs in case it can be of any help.
So far, I must have tried about 30 games, of which only one had had graphical glitches/ issues. That game is Panorama Cotton; the different background layers don't line up properly and their are some other glitches.

Soundwise, everything has been close to perfect. The only issues I've noticed so far have been with Virgin Disney games; the music just sounds off.

Keep up the good work, absolutely loving this core so far. Can't wait to see what future developments will bring. If you need a tester of some kind, please let me know.

[Sorgelig]

Many fixes in new releases.

Titan Overdrive - everything is perfect now.
Titan Overdrive 2 - mostly OK. There is unknown cause of audio lag. Arcade scene flickers and won't be fixed (as it useless for games).

Known issues:
- Outrunners USA have dots. Use Japan version - it has no such problem.
- Outrunners Japan - 2 lines are flickering in menu, otherwise game works fine.

[PhantombrainM]

I found a new bug in James Pond 3.
If the screen moves up and down, the dialog boxes like Pause menu or ingame Infos get out of screen. Previous Version is okay.

I made a video :

https://youtu.be/-Q0foiuqxlI

[Sorgelig]

I found a new bug in James Pond 3.
If the screen moves up and down, the dialog boxes like Pause menu or ingame Infos get out of screen. Previous Version is okay.

I made a video :

https://youtu.be/-Q0foiuqxlI

will be fixed in next release

[Sorgelig]

Known issue:
- Sonic: sometimes shows the garbage in level announce in NTSC mode (PAL is OK). Timing relative - not sure if it's possible to fix

[toolb0x]

There are no sound effects in Street Fighter II Champion Edition. The music is working though.

Edit: No sound effects in Super Street Fighter II either.

[cacophony]

New bug I discovered tonight: In Paperboy the houses change color as you pass

[breiztiger]

http://www.pouet.net/prod.php?which=71543

the top screen are not present ?!?

EDIT : Correct with 20181031 version thanks possible prob with scaler and 800x600 resolution i have seleted

[Sorgelig]

I found a new bug in James Pond 3.

fixed

There are no sound effects in Street Fighter II Champion Edition. The music is working though.
Edit: No sound effects in Super Street Fighter II either.

fixed

In Paperboy the houses change color as you pass

i didn't notice this problem. May be already fixed in new release.

http://www.pouet.net/prod.php?which=71543
the top screen are not present ?!?

i don't know what you mean. I've played the whole demo with only one minor issue: in one place some sprites in letters have one color different. I've compared with youtube, so it's possible the wrong color is there.

[PhantombrainM]

Thank you Sorgelig!
Im testing it in a week. Im not at home for 7 days.

[cacophony]

In Paperboy the houses change color as you pass

i didn't notice this problem. May be already fixed in new release.

You're right, it looks like the issue is fixed in the latest release. Also, the slow movement in Gauntlet 4 also seems to be fixed. Thanks!

[vanfanel]

Known issue:
- Sonic: sometimes shows the garbage in level announce in NTSC mode (PAL is OK). Timing relative - not sure if it's possible to fix

I have seen this, too. If you let the game run in demo mode, once it starts happening, it always seems to happen. It also corrupts the score, etc.

[slowmo310]

Hi, I have observed a issue with Golden Axe. When fighting Death Adder, the last boss, as soon as he uses his ground fire move, the background goes from normal colors to red. I was playing the sega 6 pack (u) Rom.

Edit: After fight with boss the background returns to normal.

[BlockABoots]

I thought the whole point of FPGA was that mimics the hardware, in this case the Genesis, so if this is the case then why do some games have graphically glitches that the real hardware doesnt?

[Sorgelig]

I thought the whole point of FPGA was that mimics the hardware, in this case the Genesis, so if this is the case then why do some games have graphically glitches that the real hardware doesnt?

Give me source code of ASICs used Genesis and it will mimic the hardware.
Do you think it happens by twist of magic wand?

[BlockABoots]

I thought the whole point of FPGA was that mimics the hardware, in this case the Genesis, so if this is the case then why do some games have graphically glitches that the real hardware doesnt?

Give me source code of ASICs used Genesis and it will mimic the hardware.
Do you think it happens by twist of magic wand?

I have no idea, thats why i was asking

So its not hardware accurate then?

[DrOG]

I thought the whole point of FPGA was that mimics the hardware, in this case the Genesis, so if this is the case then why do some games have graphically glitches that the real hardware doesnt?

Give me source code of ASICs used Genesis and it will mimic the hardware.
Do you think it happens by twist of magic wand?

I have no idea, thats why i was asking

So its not hardware accurate then?

You are very unpolite, several developers spent many hours to perfect this core! If you are not satisfied, contribute this project!!!

[BlockABoots]

Give me source code of ASICs used Genesis and it will mimic the hardware.
Do you think it happens by twist of magic wand?

I have no idea, thats why i was asking

So its not hardware accurate then?

You are very unpolite, several developers spent many hours to perfect this core! If you are not satisfied, contribute this project!!!

Im not disagreeing they have!!

I was asking a simple question!, didnt know everyone was so delicate around here.

I apologise if i offended anyone

[cacophony]

I have no idea, thats why i was asking

So its not hardware accurate then?

You are very unpolite, several developers spent many hours to perfect this core! If you are not satisfied, contribute this project!!!

Im not disagreeing they have!!

I was asking a simple question!, didnt know everyone was so delicate around here.

I apologise if i offended anyone

Reverse engineering hardware is a complicated process that takes a lot of time and dedication. For example, Kevtris spent many years perfecting his NES core. You can see some of the progress visually here if you're interested: http://kevtris.org/Projects/console/video/index.html
And as perfect as the Super Nt is there are still some remaining minor bugs that are tracked here: https://github.com/SmokeMonsterPacks/Su ... eak/issues

Sorgelig has been working tirelessly for free to get many different cores as accurate as possible, and it's been incredible to how how quickly he's managed to fix so many different issues.

[Total Eclipse]

I have no idea, thats why i was asking

So its not hardware accurate then?

You are very unpolite, several developers spent many hours to perfect this core! If you are not satisfied, contribute this project!!!

Im not disagreeing they have!!

I was asking a simple question!, didnt know everyone was so delicate around here.

I apologise if i offended anyone

Personally, I didn't think there was anything invalid in your question, however as several members here don't speak English as a first language, I think sometimes questions and comments come across other than how they were meant. Remember also that these guys write cores for fun, not profit, and do so entirely on their own time, and at their own expense.

To try and answer your question a little - you're correct, the ultimate aim of these FPGA cores is to reproduce the hardware completely. However, that's not a simple task - to give an example, even FPGA implementations of CPU's such as the 68000 aren't completely accurate, and some instructions may not take the same number of cycles to execute as a genuine CPU. It's often "good enough" for most things (the Amiga cores, for example, aren't really sensitive to the CPU instruction timings being slightly off sometimes). The same is true for other chips that make up a computer or console. Some of the functionality and timings are based on estimates and guesswork. Some of the calculations can be confirmed by looking at how emulators work for some systems, and this may at least give some kind of idea, however I think it's like the 80/20 rule. 20% of the work creating a core might give you 80% compatibility, but the bulk of the development time goes into fixing that remaining 20%.

The ST core on the MiST was good enough to run most games and productivity software, however many demos that rely on accurate instruction timings simply won't work on the MiST. Ijor has recently produced a much more accurate core for Mister, and from the comments I've seen, it's extremely accurate, allowing hardware-hitting demos to run 100%. This hasn't happened overnight though, it involves a huge amount of effort. To accurately figure out the timings required logic analysers, decapping chips (literally having the top removed and high-resolution images made of the transistor gates inside), reading through hardware reference manuals, many tests, and hours of work to convert that information into FPGA code. The results apparently speak for themselves (I've got a MiST, not a Mister, so am eagerly awaiting the port), but it's a whole load of time and expense to get to that point.

The Genesis core isn't at that level yet. It's based on what people have managed to figure out over the last 30 years, and it gets quite close, but unless someone plans to crack open the Genesis in the same way as Ijor did with the ST, people will have to make do with a core that's 99.9% accurate.

[BlockABoots]

You are very unpolite, several developers spent many hours to perfect this core! If you are not satisfied, contribute this project!!!

Im not disagreeing they have!!

I was asking a simple question!, didnt know everyone was so delicate around here.

I apologise if i offended anyone

Personally, I didn't think there was anything invalid in your question, however as several members here don't speak English as a first language, I think sometimes questions and comments come across other than how they were meant. Remember also that these guys write cores for fun, not profit, and do so entirely on their own time, and at their own expense.

To try and answer your question a little - you're correct, the ultimate aim of these FPGA cores is to reproduce the hardware completely. However, that's not a simple task - to give an example, even FPGA implementations of CPU's such as the 68000 aren't completely accurate, and some instructions may not take the same number of cycles to execute as a genuine CPU. It's often "good enough" for most things (the Amiga cores, for example, aren't really sensitive to the CPU instruction timings being slightly off sometimes). The same is true for other chips that make up a computer or console. Some of the functionality and timings are based on estimates and guesswork. Some of the calculations can be confirmed by looking at how emulators work for some systems, and this may at least give some kind of idea, however I think it's like the 80/20 rule. 20% of the work creating a core might give you 80% compatibility, but the bulk of the development time goes into fixing that remaining 20%.

The ST core on the MiST was good enough to run most games and productivity software, however many demos that rely on accurate instruction timings simply won't work on the MiST. Ijor has recently produced a much more accurate core for Mister, and from the comments I've seen, it's extremely accurate, allowing hardware-hitting demos to run 100%. This hasn't happened overnight though, it involves a huge amount of effort. To accurately figure out the timings required logic analysers, decapping chips (literally having the top removed and high-resolution images made of the transistor gates inside), reading through hardware reference manuals, many tests, and hours of work to convert that information into FPGA code. The results apparently speak for themselves (I've got a MiST, not a Mister, so am eagerly awaiting the port), but it's a whole load of time and expense to get to that point.

The Genesis core isn't at that level yet. It's based on what people have managed to figure out over the last 30 years, and it gets quite close, but unless someone plans to crack open the Genesis in the same way as Ijor did with the ST, people will have to make do with a core that's 99.9% accurate.

Yeah i guess i didnt take into consideration that Enlgish isnt everyones first language here on this forum.

Thanks for the informative reply.

So with the Mega Sg coming out soon from Analogue can we presume that that system wont be anymore compatible than the work being carried out on MiSTer currently, or will they have more in-depth research that the MiSter team have discoverer up to this point?

[BlockABoots]

Error Reporting

Game: Alien 3
Rom: Alien 3 (USA, Europe).md (No-Intro)
Issue: Game music seems to be poorly emulated and does not sound correct.