Ghosts'n Goblins on MiST

[DanyPPC]

It should beautiful if the game support VGA/YPbPr and RGB Scart output.

Many thanks guys for these improvements.

[NegSol]

This is what I get on VGA (no support for YPbPr Cable )

GnG_MiST.jpg

Have you tried with the rom file I provided?

I get the same picture with the rom file you provided. As written before, your python script for building a rom doesn't work.

I got it to work now (normal vga). The rom was the culprit. Also I build a rom for the MiSTer core with the build script by Sorgelig (based on gng.zip). While this does work on MiSTer, it does not run on MiST (even when correctly named). Just a note. So you really have to use a different roms for both machines!

Anyhow - Big thank you for the core. It is a great addition for MiST. Keep it up!

[squid4]

Looks great on RGB SCART. Thanks for this work. Can't wait for the other related classic arcade cores too!

Reds look a little brown, on RGB, but that seems to be common in many cores.

[cacophony]

jotego , thank you for the fantastic work on this! And thanks to Sorgelig for the quick port to MiSTer!

The only gameplay issue I've seen is occasionally Arthur will get stuck kneeling or as a sprite that's starting to climb a ladder (even though there's no ladder nearby). Both of these are quickly correctable by gamepad movement, and they haven't happened often enough for me to really understand the exact conditions. I'm not sure if you're looking for bugs at this point, but figured I'd mention it just in case.

Thanks again!

[DanyPPC]

Tested the last core on YPbPr cable and it's perfect !

[vebxenon]

Outstanding work!!! Thanks very much!

[jotego]

The only gameplay issue I've seen is occasionally Arthur will get stuck kneeling or as a sprite that's starting to climb a ladder (even though there's no ladder nearby). Both of these are quickly correctable by gamepad movement, and they haven't happened often enough for me to really understand the exact conditions. I'm not sure if you're looking for bugs at this point, but figured I'd mention it just in case.

I have seen it. It even happens during the demo. It happens when you press up and there is no ladder to climb. I think it is a bug in the Motorola 6809 cpu core. I have told the author of that core but he's pretty confident about his core and it is fair to say that just with that failure it is very hard to figure out what is wrong with the CPU. I still think it is the CPU because the glue logic associated to it is just too plain simple. I tried with another M6809 cpu core available but that one was even more buggy with Sir Arthur falling like the ground was just water! I have started to setup a test to compare the execution on the current CPU core with an emulator to try to find the point where both diverge. It is a bit tricky to do it, but eventually I want to get it right. So I think there will be one final update to this core in the future with that thing corrected.

One more reason I want to fix the m6809 CPU core is because it is used in the Contra arcade. I think that game can make an interesting port to FPGA too.

Outstanding work!!! Thanks very much!

Tested the last core on YPbPr cable and it's perfect !

I really love it when I see people enjoys it! Thanks a lot!

[cacophony]

I have seen it. It even happens during the demo. It happens when you press up and there is no ladder to climb. I think it is a bug in the Motorola 6809 cpu core. I have told the author of that core but he's pretty confident about his core and it is fair to say that just with that failure it is very hard to figure out what is wrong with the CPU. I still think it is the CPU because the glue logic associated to it is just too plain simple. I tried with another M6809 cpu core available but that one was even more buggy with Sir Arthur falling like the ground was just water! I have started to setup a test to compare the execution on the current CPU core with an emulator to try to find the point where both diverge. It is a bit tricky to do it, but eventually I want to get it right. So I think there will be one final update to this core in the future with that thing corrected.

One more reason I want to fix the m6809 CPU core is because it is used in the Contra arcade. I think that game can make an interesting port to FPGA too.

Nice, thanks for the info. Sounds like a tough problem but it's good to hear that you're pursing a fix.

I've been having a great time with Ghosts N Goblins the past couple days

[Sorgelig]

I think Japanese logo is way better looking They've made an art logo. English logo is too simple and dummy.

[DanyPPC]

But for most users the logo is a historical thing

[Lroby74]

Hello,
i've just donated to you 2 euro like a little prize for you awesome work with G'n'G, i think maybe i will do same for each arcade game you will convert for Mist / Mistica computers, thank you!!!!

Best Regards from Italy
Roberto

jotego, I'll donate you the price of an RGB/SCART cable - let me know where to send it. Would love to see all these cores with RGB support. Guess you're in Spain and Amigastore.eu has the cables and is in Spain.

Thanks! You can do it at my paypal address here. If I get the cable I will add support to GnG and all future CAPCOM cores (which will be quite many!)

[ReedSolomon]

someone posted code to a github today of 1943 in verilog but apparently it has no working sound. https://github.com/fredrequin/fpga_1943 not sure if anything useful there but seems to use the same sound chips as ghosts n goblins so FYI posting in case its helpful.

[cacophony]

Just curious... is the Commando hardware similar enough to Ghosts'n Goblins that you might consider doing that one as well?

[jotego]

someone posted code to a github today of 1943 in verilog but apparently it has no working sound. https://github.com/fredrequin/fpga_1943 not sure if anything useful there but seems to use the same sound chips as ghosts n goblins so FYI posting in case its helpful.

I have talked with the author and have a look at the code. He has a different approach as he has built a compatible system not a clone. He has rotated the screen 90 degrees directly on hardware. For instance, the hardware interprets the X value of sprites as Y. Then he uses a faster system bus with dual port memories, something original hardware could not do. It is a very interesting piece of work and I will probably use it as a reference. However it doesn't go in the direction I want, which is schematic/PCB based clones with the same system timings.

Hello,
i've just donated to you 2 euro like a little prize for you awesome work with G'n'G, i think maybe i will do same for each arcade game you will convert for Mist / Mistica computers, thank you!!!!

We have already talked in private but I just wanted to say thank you in public

Just curious... is the Commando hardware similar enough to Ghosts'n Goblins that you might consider doing that one as well?

Yes, that one is coming to MiST soon too.

One thing about 1942, Commando, 1943, and other so-to-speak CPS0 games is that they feature a rotate monitor. I am in favour of not rotating the screen but have the hardware draw everything as intended. If we rotate it inside the FPGA there are two options: do as Frederic in his 1943 core and generate directly a rotated image, which breaks the timing of the original system, or adding a rotation filter after generation, which adds a 1 frame delay to the output. I think that one of the nice things about FPGA is to play with original timings. If you have 1-frame delay it means that your gamepad inputs are also 1-frame delay. So for now on, I think I will not add any of these options. Maybe eventually. Probably the mister ports of these cores will get the 90º filter from day 1 as Alexey seems to be more in favour of that.

[RealLarry]

[...] I am in favour of not rotating the screen but have the hardware draw everything as intended.

That means we have to use a monitor with a pivot function? Well, I'm already using one for some of Gehstock's cores. Great arcade fealing!

[slingshot]

... or adding a rotation filter after generation, which adds a 1 frame delay to the output.

I wonder where would you get RAM for this on MiST? As it would be need to store that frame. SDRAM speed is enough?

[cacophony]

Yes, that one is coming to MiST soon too.
...

Great to hear!
Commando was a game I remember playing in the arcade as a kid and then purchasing for NES and being disappointed that it didn't look as good.
Looking forward to this one

[jotego]

[...] I am in favour of not rotating the screen but have the hardware draw everything as intended.

That means we have to use a monitor with a pivot function? Well, I'm already using one for some of Gehstock's cores. Great arcade fealing!

I love that picture. My monitor is not ready for that so I am forcing it a bit. Still, looks great to the screen on a tilted monitor.

... or adding a rotation filter after generation, which adds a 1 frame delay to the output.

I wonder where would you get RAM for this on MiST? As it would be need to store that frame. SDRAM speed is enough?

Not enough RAM inside the FPGA. The screen is basically 256x256x12 bit. I think I have a couple of free slots in the SDRAM yet. Or maybe I can move the refresh cycle to the blank period only. I will have to think about it.

[slingshot]

Not enough RAM inside the FPGA. The screen is basically 256x256x12 bit. I think I have a couple of free slots in the SDRAM yet. Or maybe I can move the refresh cycle to the blank period only. I will have to think about it.

I think about burst write/read to/from a line buffer in BRAM. But probably would steal too many RAM cycles from the CPU.

[Sorgelig]

Not enough RAM inside the FPGA. The screen is basically 256x256x12 bit. I think I have a couple of free slots in the SDRAM yet. Or maybe I can move the refresh cycle to the blank period only. I will have to think about it.

with correct SDRAM addressing, frame buffer access can be used as a refresh.

[jotego]

I have updated the core fixing the imaginary-staircase bug, which occured when you hit the up/right diagonal and the character would start climbing in the air. I have also added keyboard support.

Please find it here.

[DanyPPC]

Thanks Jotego for this update

[LarsDragon]

Hi Jotego, the update core don't show image over 15 khz scart connection, only menu if you press F12. The rom is correct, worked in previous release.
In 1942 core that happen too.

[slingshot]

Seems game_{r,g,b} signals are not connected to jtgng_mist_base.

[jotego]

Seems game_{r,g,b} signals are not connected to jtgng_mist_base.

OMG! That is probably the cause. I have fixed it (and pushed it into the repository) but I have not been able to test it yet. Thanks a lot!