Arcade Simulator - Layout Permissions

[serene02]

I’d say it’s probably more like an SWP, it will be a compensated machine with very little real skill involved. 

Massive thanks to JP though, what a total legend getting this working, I’m max skillz man!!  Always nice to see new possible emulations, I’m sure some will be interested in this.

J

[johnparker007]

5 hours ago, Zippy said:

I fail to see what excitement you can get from this, reminds me of a fruit machine with no hold/nudge buttons. If this is pure skill and you are skilled enough you could empty it but what happens % wise? If you pressed the button all they way every time for a £100 of credits would it still pay (in other words if a complete plank played it and lost would it force itself to pay)? If it is a true skill machine then it wouldn't.

Also if this could play in MFME you'd need a cab, a space bar ain't gonna cut it.

Yeah I think it'll make more sense when someone makes a layout and figures out that middle button, plus any dips/invert button to disable 'autoplay'.  Then it appears that when you hold down the button, the flashing tool moves up the middle column, at a random constant speed.  Then there is some level of inertia when you release the button, based on the speed, before the flashing tool comes to a halt.  And the aim is to get multiple to stop on the same row, then you get a prize and I think it resets.  You can also choose how much to keep staking as zero credits also resets, so I guess that is another part of the skill.

At least that's what I got from it anyway!  I think it'd play fine on space bar as it's just a standard button.  Though whether you'd have much fun playing this for a prolonged period remains to be seen  

It doesn't look great from first impressions to be fair, but more likely someone's hobby attempt to make a skill based machine in an existing shell, but hey that's how those companies like Bwb and Pcp did business 

The company Jabro Games, appears to have been a hobby that barely got going -as their website, it's still there, but it's just an unfinished Wordpress template:

https://www.jabrogames.com/

And on the youtube video linked above, the official company phone line is just a mobile number, which means no dedicated office.  It's like if I made and sold a fruit machine in my spare time I think  

Edited May 14, 2022 by johnparker007

[johnparker007]

2 hours ago, serene02 said:

I’d say it’s probably more like an SWP, it will be a compensated machine with very little real skill involved. 

Massive thanks to JP though, what a total legend getting this working, I’m max skillz man!!  Always nice to see new possible emulations, I’m sure some will be interested in this.

J

Ah just saw this on page two  - from the official Jabro youtube video, it says:

Skill Shuffle SWP Game by Jabro Games.

So sounds like you're right then, it's not 100% skill (compared to the 'Stop The Queen' game).  So it probably fiddles what skill stops are possible or something  

Oh my god, I just checked their other videos, they made more obscure machines inside other cabinets!   



Check out Motorbike Grand Prix - it's blatantly been wedged into a Horizon cab!    He'll use whatever he's got to hand I guess, fair play  

As a side note - if we have the ROMs for these other games and they are MPU4 (seems fairly likely it's be the same tech if Jabro is a one man band), I'll very likely be able to quickly remove protection (as I know his protection routines now) so they can be played via emulation.

Did some further digging, they did make, or at least produce some more modern games later on:

So god know what's going on with them lol  

Edited May 14, 2022 by johnparker007

[spa]

2 hours ago, serene02 said:

I’d say it’s probably more like an SWP, it will be a compensated machine with very little real skill involved. 

Massive thanks to JP though, what a total legend getting this working, I’m max skillz man!!  Always nice to see new possible emulations, I’m sure some will be interested in this.

J

Well, would be interesting to see how low it can be set as rumors are, it's less than 50%!

[serene02]

You’d be able to find out in mfme, test mode etc.  might show % and actual %?

J

[andrew96]

Thanks for the tag, yes interesting for sure! I still cannot make out why the real chr chip sends out 3 different figures for the same byte sent to it when interrogated 3 times in a row, where as a different figure that's constant is used for mfme and it takes that as correct..... but there definitely is something in using the real chr chip when it reads the different figures which makes it payout the correct percentage, and when patched with the single figure the percentage goes out the window!

Patching is definitely good for mfme as it allows you to play the game! but in a real machine..... well I definitely would not want to play one in a arcade that's been patched like that! worse than the fairground traveling fairs rom hacks!!

It still needs working on for sure to work in the 3 different values returned in the lamp table for the same byte! barcrest did a good job in protecting it all in a good few ways which we have not yet hacked entirely!

[johnparker007]

14 hours ago, andrew96 said:

Thanks for the tag, yes interesting for sure! I still cannot make out why the real chr chip sends out 3 different figures for the same byte sent to it when interrogated 3 times in a row, where as a different figure that's constant is used for mfme and it takes that as correct..... but there definitely is something in using the real chr chip when it reads the different figures which makes it payout the correct percentage, and when patched with the single figure the percentage goes out the window!

Patching is definitely good for mfme as it allows you to play the game! but in a real machine..... well I definitely would not want to play one in a arcade that's been patched like that! worse than the fairground traveling fairs rom hacks!!

It still needs working on for sure to work in the 3 different values returned in the lamp table for the same byte! barcrest did a good job in protecting it all in a good few ways which we have not yet hacked entirely!

Yes I remember your 3x value findings in comparison to what we see in MFME, all was very surprising to me... as there is stuff like a routine in the rom that has a copy of the table (initial 64 values), and I think it just does something like:

write byte => $0800
read byte <- $0800
compare to byte in ROM table
(jump to 'Security Error' if not equal)
increment ROM table address pointer
loop until 64
...fall thru to security check passed...

It doesn't have any special code for processing the 3x values, or performing the 3x read/writes... all very odd!  I don't have a real machine here (or the time sadly) to do more of a deep dive into how the heck it's working, but it is all super confusing to me for sure! I don't see it doing the 3x cycles per value... 

When people leave MPU4 Chr-protected machines running in MFME via 'Autoplay' (where it slowly plays through thousands of credits), don't they end up coming out at roughly the correct In/Out percentage ratio?  This info in the MFME Game Manager (rightmost three values are In/Out/%):


I thought they generally worked out around the correct % ratio over time... but I've not really tested this I'll be honest!  

"but there definitely is something in using the real chr chip when it reads the different figures which makes it payout the correct percentage, and when patched with the single figure the percentage goes out the window"

Do you have the name of an example MPU4 machine that you believe does this please?  Then we could put it in Autoplay in MFME, reset its statistics and RAM, and leave it playing for say 24 hours solid (running its original, unmodified ROM code).  And we can then see if this bears out when properly tested.  So after say £10,000 in, has it paid out the expected %? 

If it is a provable, repeatable problem (that running certain MPU4 machines under current emulation techniques results in broken % payouts over time), maybe in MAME one day we could develop a proper emulation (using this 3x value per write/read model), it's just a few simple routines in the mpu4 c driver that handle Chr emulation, nothing very complex - I know them pretty well now from learning to hack out the routines in the 6809 ROMs  

Edited May 15, 2022 by johnparker007

[andrew96]

I have heard it happens but cannot give you any more advice on what machines ( I been waiting for one digital pic of the top door switch arrangement on a red gaming cab now since the start of the red dwarf project, and after asking the arcade owner lots of times I am still waiting years later.... so not a answer on what machines work incorrectly with fixed lamp table I am going to get I think!!)

the chr table is fixed and that's not the issue, that is not too much of a hassle to overcome, its the 3 x scan of the lamp table with different figures that is the head scratcher!!  The emulator is just that, it emulates! it is not the same as running real hardware as we know!

More work needs to be done to recreate the 3 different tables used.. but don't think that's going to happen as MFME is now fixed with just one set of figures and the author is not about to look into this or change it, Just a heads up if using a patched lamp table in the rom its not likely to play correctly in a real machine... I can't be any more help than that really, MFME is beyond me.

[johnparker007]

1 hour ago, andrew96 said:

I have heard it happens but cannot give you any more advice on what machines ( I been waiting for one digital pic of the top door switch arrangement on a red gaming cab now since the start of the red dwarf project, and after asking the arcade owner lots of times I am still waiting years later.... so not a answer on what machines work incorrectly with fixed lamp table I am going to get I think!!)

the chr table is fixed and that's not the issue, that is not too much of a hassle to overcome, its the 3 x scan of the lamp table with different figures that is the head scratcher!!  The emulator is just that, it emulates! it is not the same as running real hardware as we know!

More work needs to be done to recreate the 3 different tables used.. but don't think that's going to happen as MFME is now fixed with just one set of figures and the author is not about to look into this or change it, Just a heads up if using a patched lamp table in the rom its not likely to play correctly in a real machine... I can't be any more help than that really, MFME is beyond me.

Ah ok, thanks for the reply   

So to be super clear in case I've got the wrong end of the stick! - you are saying that: some MPU4 games running under MFME have broken payout percentages?

 

Edited May 15, 2022 by johnparker007

[andrew96]

nope, I am saying in REAL machines the patching of the use of a lamp table (single number instead of 3 used in the real chr chip lamp table) causes issues in real machines, I dont know about MFME as I only have interest in real machine stuff

[johnparker007]

22 minutes ago, andrew96 said:

nope, I am saying in REAL machines the patching of the use of a lamp table (single number instead of 3 used in the real chr chip lamp table) causes issues in real machines, I dont know about MFME as I only have interest in real machine stuff

But unfortunately we currently have no idea which machine(s) then, ok...  If ever you get more info, please do let me know as it would be good to investigate if this does actually happen   Currently we have zero info of an actual machine with a problem, so I can't really progress any solutions etc.

But would definitely be keen to look into it if it actually happens, always looking to improve stuff!  

Edited May 15, 2022 by johnparker007

[serene02]

This was an issue with BWB, in fact Chris fixed a machine that was causing the machine to play incorrectly via the chr, I believe it was a BWB Bg Match, or Cup Final, incidentally I think it was specifically BWB machines that he had the issues with.

I think it was Chopley who raised the issue with Chris, who then did fix it.  But it will still exhibit the same issue in older releases of mfme I imagine.

Might be worth looking for the thread that Choppers created.  Think it was BWB Cup Final though.  The gameplay was definitely affected

J

[serene02]

We have issues with certain games that play differently under Epoch tech.  In particular a machine called Simpsons (blue).  In the emulator it never reaches a point were it offers the streak, which it did in the wild?

A fair few strange quirks.  I do remember that Chris said that the tech weren’t fully emulated, just enough to get the machines working. I don’t know how true this is in later revisions or if this could cause these strange quirks?

J

Edited May 15, 2022 by serene02

[Chopaholic]

17 minutes ago, serene02 said:

This was an issue with BWB, in fact Chris fixed a machine that was causing the machine to play incorrectly via the chr, I believe it was a BWB Bg Match, or Cup Final, incidentally I think it was specifically BWB machines that he had the issues with.

I think it was Chopley who raised the issue with Chris, who then did fix it.  But it will still exhibit the same issue in older releases of mfme I imagine.

Might be worth looking for the thread that Choppers created.  Think it was BWB Cup Final though.  The gameplay was definitely affected

J

It was Cup Final, it was one of my FME Most Wanted Machines, but it just wasn't playing right. I made super sure of it before posting because Chris didn't suffer fools gladly, but I was convinced enough to say to him the machine wasn't working properly, and he tweaked something in MFME to make it work properly. (CHR related.)

I'll have a search and see if I can find the thread.

Chris was always so good at fixing anything and everything he could, R.I.P. fella, one of a kind.

[serene02]

I echo your comments about Chris Al, he was always super cool if anything wasn’t working correctly.  He would definitely tell you straight, which I admired .

I’m in no way trying to taint his superb work with mfme.

J

[johnparker007]

18 minutes ago, serene02 said:

This was an issue with BWB, in fact Chris fixed a machine that was causing the machine to play incorrectly via the chr, I believe it was a BWB Bg Match, or Cup Final, incidentally I think it was specifically BWB machines that he had the issues with.

I think it was Chopley who raised the issue with Chris, who then did fix it.  But it will still exhibit the same issue in older releases of mfme I imagine.

Might be worth looking for the thread that Choppers created.  Think it was BWB Cup Final though.  The gameplay was definitely affected

J

Ah right, so it's certain Bwbs - yeah some of them definitely have different stuff going on the with Chr routines.  Right so it sounds like standard Barcrest MPU4s generally have the correct payout percentage in MFME then.  Additionally, if the Chr fix from Chris worked for the Bwb Cup Final machine, and payout was then correct - it was not caused by reading a lamp table vs doing a theoretical 3x read (we don't see that in the 6809 ROM code, but we do I think when Andrew was probing the Chr chip itself with his fluke meter device out of circuit).

Good to know!   One day it'd be good to get a test rig, or a real MPU4 machine here, so I can really probe further... perhaps there's a debugging port on the MPU4 board or something to allow us to actually step instructions etc on real hardware...

11 minutes ago, serene02 said:

We have issues with certain games that play differently under Epoch tech.  In particular a machine called Simpsons (blue).  In the emulator it never reaches a point were it offers the streak, which it did in the wild?

A fair few strange quirks.  I do remember that Chris said that the tech weren’t fully emulated, just enough to get the machines working. I don’t know how true this is in later revisions or if this could cause these strange quirks?

J

Ah right, I think there's just a skeleton driver for Epoch fruit emulation in MAME, so tons to do there if anyone gets to re-reverse engineering it!  If Chris just did enough to get the games running, there definitely could be problems with the emulation accuracy in MFME (program bugs etc).

[johnparker007]

4 minutes ago, serene02 said:

I echo your comments about Chris Al, he was always super cool if anything wasn’t working correctly.  He would definitely tell you straight, which I admired .

I’m in no way trying to taint his superb work with mfme.

J

Yes, the up to date version of MFME is great  - for instance the patch I developed on page one of this thread for the obscure 'Jabro Games Skill Shuffle' machine, would've taken far longer to create without it.  The vast majority of people I'm sure don't use this feature at all, it's this debugger and it is really handy, for debugging and even changing ROM code as it executes:



Brilliant work from Chris  

[serene02]

Chris did mention that the real hardware used some fake read writes on the CHR.  I can’t confirm any of this though, but maybe it accounts for those 3 x reads?  Just put there to trick any would be hacker?

J

 

[johnparker007]

14 minutes ago, serene02 said:

Chris did mention that the real hardware used some fake read writes on the CHR.  I can’t confirm any of this though, but maybe it accounts for those 3 x reads?  Just put there to trick any would be hacker?

J

 

They could be peppered around the startup code etc, to stop hackers like you say... it does make things more difficult if there's dummy Chr reads... as you can then spend time hacking those out, to find they didn't do anything!

I just played some brief test credits on the Cup Final layout from the legacy section, doesn't seem to be giving me 100% payout.  The lamp columns are all zero, so they probably don't even factor in for Bwb Chr system.

I did a brief test: some emulation timings seem broken, and some samples are playing back too fast.  I don't think it has any 100% payout issue though, just some emulation bugs (that can be fixed in MAME at some point).

The broken sample playback (too fast) can be heard around 4:43 as I play the feature board:

 

[serene02]

You’d have to play it in an older version of mfme JP.  Maybe 18.1?

J

[johnparker007]

37 minutes ago, serene02 said:

You’d have to play it in an older version of mfme JP.  Maybe 18.1?

J

Ah thanks, though we don't need to investigate the older broken version of the emulator, for the Chr stuff (unless you mean the sound isn't broken on older versions?) - the most up-to-date one is showing that it can work (and doesn't have issues with the hypothetical 3x read thing, which isn't a thing in terms of what's visible to the CPU IO - just something Andrew found when probing the Chr chip out of circuit).

Here's an example of a standard Barcrest reading lamps on the fly.  It just sends a 0 to reset the Chr to lamp addressing, another value to address it (which is convoluted), and then a read.  It doesn't sequentially read 3x values, this is not really a thing I don't believe, as the ROM code is there for us to read and observe with breakpoints.  Doesn't matter if real machine or emulation, the 6809 code in ROM is what the CPU executes, as it stands.

No magic going on there honestly   It pokes register A contents the Chr chip, then pokes register B contents to the Chr chip, then peeks the result back into register B.  That will be exactly the same on real hardware, it's just a 6809 chip and ROM code to consider  


One day perhaps I can buy an MPU4 machine that's really unpopular, so it's cheap!  Like a really crappy standard hardware Barcrest that no one likes    Then I can use that as a test jig - and prove via the in/out meters that any given game (that I flash in) leaves the meters in correct %age after a few hours of manual play (even with some future form of my 'stealth' Chr free patches I have planned in place).  All straightforward really.. though a fruit machine takes up a bit of space I don't really have lol 

Edit: just had a look, there are cheap'ish' ones to be had - an MPU4 Action Bank went for £250 a couple of months back, as the glass artwork is all cracked and peeling... but since I'd be using it as an MPU test jig for testing new Chr patch development, it wouldn't really matter  

https://www.ebay.co.uk/itm/fruit-machines-coin-operated-gaming-/334347684022?hash=item4dd8acc8b6%3Ag%3AatkAAOSwQGFiHgEz&nma=true&si=NdObeAXiSbbnkDnsr6RFlzps9TI%3D&orig_cvip=true&nordt=true&rt=nc&_trksid=p2047675.l2557

Not something I've got time to work on these days with AS, but at least it's possible to get a cheap one to use for testing  

Edited May 15, 2022 by johnparker007

[Zippy]

Sounds like me and Tony after a few beers and some chasers

[serene02]

@johnparker007You should be able to get an MPU4 much cheaper.   Someone might even have a test rig we could club in a buy for you bud?  They take up less space I would have thought     Andrew might even have one.

J

[johnparker007]

5 minutes ago, serene02 said:

@johnparker007You should be able to get an MPU4 much cheaper.   Someone might even have a test rig we could club in a buy for you bud?  They take up less space I would have thought     Andrew might even have one.

J

Well that would be very kind of folks but it's not something I need at the moment, and not for Arcade Sim anyway - the patched Chr chips I make seem to work 'ok', though I can one day do better patches I think given some time to develop the new technique ('stealth' rewrites of checksum and chr+lamp code, rather than the standard common approach). 

It's more so that I could fix the issue Andrew was mentioning about payout percentages - as I believe if I had even a test jig, and then bought some reels for it, I could run lots of different MPU4 games in it.  So then I could make a stealth patched ROM, play it in the rig (arkward but doable), and then check the ratio between the in/out meters after a few hours of manual play to check it wasn't paying out 100% (or more) of what went in.  I don't know what machine(s) this is on though, perhaps is more to do with the Bwb Cup Final and similar mentioned above, rather than affecting any Barcrest machines... it may turn out that the new patches are fine anyway!    Just that Andrew has been saying about these problems wrt patched ROMs (ie. not Bwb but Barcrest), but unfortunately we don't know which games.

[andrew96]

its difficult to get any info on anything, cant even get a picture of behind a lock showing the position of the switch that mine doesn't have!