David Haywood's Homepage
MAME work and other stuff
April 30, 2013 Haze Categories: General News. 50 Comments on UME 0.148u4 + What’s New in 0.148u4

UME (logo by JackC)

UME (Universal Machine Emulator) combines the features of MAME and MESS into a single multi-purpose emulator. The project represents a natural course of development for the emulators which already share large amounts of code and is part of an ongoing effort to unify development efforts and provide a single emulation platform for users and developers alike.

As an end user this means that the software provided here is not only capable of emulating arcade machines like the baseline versions of MAME, but in addition can emulate a large number of home computers and consoles from across the world using the very same code, developed by the very same team of developers.

0.148u4 Windows binaries (32-bit and 64-bit) (Self Extracting 7-zip) (all MAME / MESS tools included, both 32-bit and 64-bit versions in tool32/tools64)

The source is identical to that found on mamedev.org (SVN revision SVN 22616 / 0.148u4)

Non-UME binaries

In addition to providing the UME binaries I’ve also included a package with the individual legacy MAME/MESS executables here, personally I prefer the everything under one exe UME solution but I’ve noticed it’s not always easy to find binaries of the regular u builds with them not being offered from the official site so this is my attempt to address that.

Latest U release binaries for UME, as well as MAME & MESS can also always be found on the page linked in the box on the left
These binaries are coming from Mamedev (me) so are as official as you’re going to get for a u update.

Please note, if you want to compile a linux build you should check the source out from the official SVN instead of downloading it, there is an issue with some source files and line-endings whereby the linux python scripts don’t agree with the Windows style line endings you get on the downloaded source. Details of the SVN server are still on mamedev.org

What’s New

You can read the various whatsnew files on mamedev.org
From MAME, From MESS

Points of Interest

0.148u4 like 0.148u3 before it isn’t the most interesting of releases, but I’ll cover the key points shortly

In MAME terms there have been improvements to several of the video fruit machines, the ROMs were dumped from a Sppoky Night board, correcting the graphics and sounds in it, although the board found was an original ‘1st revision’ board, and it’s possible the graphic roms on a 2nd edition should differ. Official promotional material shows a correct spelling of the word ‘POINT’ in the bottom left corner of the playfield, this comes from the graphic roms and is definitely incorrect in the ones from the board that was located.

Spooky Night Spooky Night
Spooky Night Spooky Night
(Spooky Night now works in 0.148u4)

Along similar lines is the work done by Robbie on Good Luck 2, a simple but well presented poker game using rotating turtle graphics by default to present the card value, you can see more pictures on his blog, as well as ones of Mondial which will almost certainly feature in the next release.

Good Luck 2 Good Luck 2
Good Luck 2 Good Luck 2
(Good Luck 2, another poker variant now working in 0.148u4)

Maybe slightly more interesting is that a supposedly unprotected version of the IGS PGM title Oriental Legend Special/Super was uncovered by XingXing. I’m not quite sure of the story behind this one, an unprotected IGS game seems unlikely, but this cartridge was said to be licensed to or produced by TenCent. I’d suspect it to be a bootleg, but it appears to be a unique build of the game with a 2003 build date, for the China region only. It’s marked as Working in MAME, although the high score table doesn’t appear properly (adding further to the mystery because that’s one of the protected elements of the original game, although it could be an emulation glitch) People have reported that MAME sometimes crashes running this, but I’ve not experienced that myself, if it crashes for you and you can get a backtrace please post it, it will be invaluable.

Oriental Legend Super Oriental Legend Super
(An unprotected? version of Oriental Legend Super turned up)

The new Ball Boy clone of Snow Bros 3 is also an interesting case, the undumped database shows Ball Boy as a 2001 game, with a different title screen to the dumped version however the dumped version bears a 2003 copyright and is clearly a hack of the existing Snow Bros 3 set (the ending still shows Thank You Snow Brothers)

I have a feeling the original hack was a 2001 release, called Ball Boy, released in Korea (it makes sense, the game has footballs instead of snowballs and was done to coincide with the 2002 World Cup in Korea / Japan) From that point the game was likely hacked by Syrmex for Mexico, restoring the Snow Bros name (slightly ironic if the original Ball Boy name was meant to hide that it’s just a hack of the original Snow Bros) after that for some reason it then got hacked back into ‘Ball Boy’ by another bootlegger. All versions appear to have the protected Sound MCU tho, so this new clone doesn’t really improve the sound emulation which still hinges off a rough simulation. This new Ball Boy set also has a crudely hacked in World Map screen, but the backgrounds used ingame have little (nothing..) to do with the locations on the map! To make matters even stranger the World Map (not present in the Mexican released Snow Bros 3 set) shows Mexico as the final destination. The story of Korean hacks is always something of a trainwreck.

Ball Boy Ball Boy
Ball Boy Ball Boy
(This version of Ball Boy appears to be a hack of the Snow Bros 3 set, but there should probably be an earlier one predating it too)

One other significant clone to add to the ‘against all odds’ list is the 4-player Japanese revision of DD Crew, it’s still a terrible game, but always good to see working FD1094 based games show up because we need them in working condition for the keys to be read out and every single one is well past any expected shelf life right now due to the suicide batteries.

DD Crew (Japan 4 player) DD Crew (Japan 4 player)
DD Crew (Japan 4 player) DD Crew (Japan 4 player)
(A working 4-player Japanese DD Crew PCB turned up too)

None of this is too exciting, even with the additional trivia behind each of the sets and there were a number of other sets added where I really have nothing of note to write about them at all, nothing really ground-breaking anywhere. The swapping around of the puckman sets is probably going to irritate many users, but it’s long been suspected that the main set was a hack due to the protection being patched out, so u4 corrects that by swapping them around.

Cross-project work has seen continued improvements to the Playstation emulation, with reworking of the CD-ROM code from Carl and general fixes to various modes of operation from smf, mostly of benefit to the actual Playstation emulation, and mentioned with a word of caution because the code is still very much in flux but definitely heading in the right direction. The best place to follow that progress is on the bannister.org forums, also not a bad place to discuss MAME and emulation in general given the continued decline in standards over at Mameworld. (It wouldn’t surprise me if we see some more official form of forum soon simply to disassociate Mamedev with what’s going on there)

On the MESS side judge has started to look a little at the BBC drivers because they’ve been decaying for a while, and things like the cassette loading failed to survive several key core changes and device updates (the original device implementations it used had system specific hacks and that broke once it was migrated to common code etc.) Cassette loading is said to work again, although there are still some issues with saving, so things probably aren’t quite right yet, it’s a good example of where MESS would really benefit from having experts associated with each system to help maintain and develop drivers for systems they know well, however in many cases such experts are lacking. Given the popularity of the Acorn machines in schools across the UK that actually surprises me for these systems, it amazes me that the Acorn Archimedes emulation still does almost nothing too I would have expected a large number of people to be well versed in the systems when you look at how long they were around for.

etabeta also continued with his slot work, and general improvements to the NES emulation, adding support / fixing up various mappers and the like, that constitutes most of what was changed in 0.148u4 for MESS. I might pick out some specific examples, but I don’t think there’s anything really groundbreaking there either even if it is important work. The Nantettatte!! Baseball lock-on cart support is a neat little feature tho, and shows it wasn’t just Sega experimenting with such technology.

The Z800x fixes for the M20 system are good to see, with the work done by Christian Groessler showing what I mean about the benefits of having an expert associated with certain systems, it’s unlikely many other people care about / understand that system and CPU family well enough to be doing the work he is doing, which means the contributions pleasing to see because such dedication can easily make MESS the best in field emulator for any given system and that’s where it needs to be if possible.

Looking back at MAME there were also skeleton drivers for the Pinball 2000 unit, and newer Golden Tee Fore! systems added, and while both (or at least the latter) will certainly interest a few parties it remains to be seen if there will be any further immediate progress, skeletons are often placed in the source simply to ensure information isn’t lost for potential work much further down the line. There were some regression fixes too, tidying up after recent core changes (for example Turbo Sub was fixed to work with the more accurate CPU cores) but the lack of a working MameTesters site is proving to be something of a hindrance there.

Like I mentioned at the start, nothing too exciting this time, although still enough changes to make it a worthwhile update if things like the Playstation emulation interest you because the significance of the changes made there should not be underestimated.


You can follow any responses to this entry through the RSS 2.0 feed.

A unprotected bootleg of oriental legend is an interesting find
i wonder if there are others for some of the pgm games that
dont work as yet in mame since sorting the protection is so difficult
for those games a unlicenced bootleg might be the best chance
of playing gladiator road of the sword in mame anytime soon

I’m not sure it is a bootleg, it might just be there was demand for a version at a time when IGS had none of the original (older) protection components available, and hadn’t updated the protection code to work with newer ones (and once they had they released it as Oriental Legend Super Plus) The build dates and version number put it in an odd position.

I find it unlikely that there will be an unprotected version of The Gladiator or any of the later games because they rely heavily on the ARM (almost the entire game runs on the protection CPU, the 68k alone is too weak)

For those that don’t know/remember the BBC Acorn (this is more for younger MAME users and those ‘across the pond’), to give an example how mainstream these computers were in school these where coming into schools in 1981.

A school then to be at the forefront of IT had one of these machines. By 1989 many secondary schools around the UK had about 20-30 in a computer room.

These were then replaced slowly but surely by IBM clones running windows 3.11, but some who wanted to be a bit maverick also ran apple based machines!

For a certain generation, these computers were our first taste of computers, not only that BASIC programming language, unfortunately this is where my programming begins and ends, but thankfully it gave some 5 year olds ‘back in the day’ enough experience to write drivers for an emulator!

30 years later! (it made it sound that a 5 year old could write a driver… sorry for that my bad.

The science department at my school had BBC Master / Micros hooked up to all sorts of sensors / test software / electronics software etc. for many, many years even after all the Acorn Archimedes systems had (stupidly) been replaced with vastly inferior Windows PCs (how anybody thought that was an upgrade is beyond my understanding)

The Acorn systems were quite something, ahead of their time in terms of functionality and software ideas to say the least.

I seem to remember there was a word processing package on one of the ROM chips of the BBC micro, does anyone remember the command to get this up? I remember it having things like centre/left/right, as well as find/replace. But I do recall it didn’t have many fonts!

We were discouraged to use it for coursework as it was considered ‘lazy’ to word process…nowadays the same teachers demand stuff to be done on word…so they can read the handwriting!! (so by my logic – teachers are lazy!)

might have been an optional expansion rom? they were very flexible systems and obviously having rom expansions for such things was very practical / useful.

unfortunately I don’t know much more about them because they were the systems we used in schools rather than the ones I had at home. I do remember lots of custom keyboard overlays with details like you mention on however, but it’s possible they were part of said package (it’s been too long)

Was the word processor Wordwise? That seems to have come on a ROM chip. It looks like typing *WORDWISE is supposed to start it up (assuming the ROM is loaded).

concerning NES… well, the Nintendo World Championship 1990 cart is interesting (it was the last official PCB completely unsupported in MESS) and the proper Game Genie emulation is funny

also, if you compare Klax, Road Runner, Alien Syndrome, Family Circuit 91, Impossible Mission II, Napoleon Senki, Sachen’s Princess Maker or Mi Hun Che between u3 or u4, you will see what the progresses have brought

the “problem” is that we already emulate well ~85% of the titles, and the rest needs a PPU rewrite to get fixed, so it’s difficult to bring anything ground-breaking in NES… the big progresses happened in 2009 ;)

well I mean groundbreaking vs. what’s already out there as well, I don’t believe MESS is actually breaking any new ground with these (figuring previously unknown things out) or am I wrong?

That’s what I really consider progress in the projects, where things are being done that nobody has ever done before, and previously unknown facts being discovered and documented.

Retracing the steps of others is fine, and obviously needed to bring MAME / MESS on par with some other emulators (and thus make it a viable alternative with no real drawbacks) but the real value of the project is as a vehicle for discovering and documenting new things, aiding others etc.

When I say groundbreaking I typically mean that (and as emulators they’re fairly well positioned to do such things, especially when it comes to emulation of unique expansion cards, and previously uncared for pirate carts etc.)

Of course things like the Spooky Night progress in MAME cover previously uncharted territory too, but I don’t really consider them groundbreaking because the actual work was done long ago, they just work given the correct roms. Proper emulation of something like the speech chip used on Gorf would be ‘groundbreaking’ IMHO, because the only things to emulate it so far simply rely on crude hacks (lots of samples stuck together) and sound nothing like the original.

Not to knock the progress, I just don’t really feel there is much that falls into that category as of the last few updates.

I’d still consider the main project as a UME style thing more groundbreaking because it signals real intent and direction, although even that really is just presenting the same things in a different way too, much like most of the progress of late ;-)

This is understandable, there is only so much ‘groundbreaking’ work left to do, and I guess you could argue that while the slots stuff doesn’t cover new ground it does cover new ideas, and makes MESS a more professional emulator, much as say an SH4 recompiler and 3D card accelerated software rendered for Naomi wouldn’t bring new playable games that nobody has played elsewhere, but would show a very significant improvement in the project, and better prepare it for other work where maybe new discoveries really can be made. MAME got where it is by making a difference, and really needs to continue to do that, challenge peoples perception of what is possible, and where the project can go. MESS needs to do the same too.

The downside is of course with so few interesting things left it might actually make more sense to space them out a bit, if we emulated Raiden 2 tomorrow and I put up a UME build with it then I’m sure a lot of people would upgrade to that build (regardless of whatever else was in it) then never, ever update again ;-)

The Playstation progress strikes me as interesting because it seems that even with a number of ‘mature’ emulators out there many things aren’t fully understood, or at least aren’t documented properly in any open project; pSX for all the acclaim it gets from certain developers actually suffers from many of the same problems as MESS did before the recent progress, and while people say some other emulators do handle things better it doesn’t feel like anybody is 100% certain over certain aspects of the system emulation right now.

then for NES there is even less hope for anything ground-breaking ever to happen
basically NEStopia contains accurate emulation of 99% of what has ever existed for NES

the only new achievement I can think of about NES is real emulation of the ADPCM chip in Jaleco sport games (and in a Bandai game), but that would require probably a decapping unless OG’s efforts on a similar chip produce any result *and* the same approach happens to work on the Jaleco and Bandai ones…

last summer, a large bunch of very rare dumps got released, but it turned out all of them were running on common hardware, so they just fit the existing emulation
same happened with the very first pirate game (Super Maruo) whose dump surfaced last fall

so what is left? we could maybe improve the emulation of the Datach barcode reader (which NEStopia already partially emulates anyway) or we could reverse engineer the Famicom tapes (which NEStopia only supports as a sort of snapshot format), but they would probably go unnoticed and qualify more or less as ground breaking as Nantettatte!! Baseball lock-on ;)

other than that, the best we could do is to stay on par with Cah4e3’s new dumps and the emulation of the corresponding pcbs, and to fix the remaining issues we have

It sounds funny, but groundbreaking would be simple save state support for NES. Many NES games, which required keeping an NES powered on for days/weeks/months to finish back in the day, are simply unplayable in MESS without it. Regardless of accuracy, NEStopia and others are better options for this reason. Some of the slot stuff is pretty interesting to watch, but without save states it’s mainly just a curiosity.

On the subject of PS1, Mednafen is far better than even pSX and is fully open-source. Excepting a few VERY rare cases, it’s basically perfect, supports cross-platform rumble and other nice features, and handles a lot of stuff pSX can’t touch.

I’ll have to give Mednafen a spin one day. I guess maybe one of the reasons some MAME developers don’t hype it up as much is because of the GPL license, which we’ve always preferred to avoid for MAME as to discourage people from packaging it up and selling it without permission, thus making the code of no use to MAME.

Proper working Save States in NES would be great yeah, I’m not really sure what the issue is to be honest, adding Save support for anything in MAME is an easy task these days. I wouldn’t call it groundbreaking tho, other emulators already have more advanced support with all the rewind options which is beyond any kind of save state logic MAME/MESS have right now ;-) It would however help with the popularity of the NES driver in MESS, I’m sure about that.

eta confirms pretty much what I was thinking with the NES tho, there’s nothing left anywhere that somebody hasn’t already done / discovered, even the ‘enhanced’ NES clones have been covered (where the clone PPUs don’t adhere to the limits of the originals etc. much like a lot of modern arcade bootleg hardware)

It would be nice to say one day that MESS is the only NES emulator you’ll ever need for any normal or debugging use (even if it lacks ‘rewind’ style frills) but NES emulation has been done to death to the point where there is little novel or interesting to highlight, there must be literally hundreds of NES emulators on par with MESS. Articles about the lost NES gems tend to be quite interesting tho, with such a vast software library there are many unappreciated titles but that’s not really the focus here.

I guess it would be quite cool to see those Super Mario Bros. arcade bootlegs running, where they appear to have bolted a Z80 + other extra hardware on to replace something (the sound system?) they’re surely unique and could probably be made to work with a bit of comparison work between the original versions and the bootlegs (suprmriobl, suprmriobl2)

Good points. Mednafen is indeed GPL unfortunately, but perhaps the MESSdev could inquire about an exemption is MESS’ case (hasn’t that been done before?) Obviously that would depend on how the Mednafen dev feels about that, but it might be worth a shot.

I’ve also been a bit surprised save states haven’t been added after all this time (been wondering for several years actually), perhaps it’s had something to do with the stability of the driver but it seems a good time at this point. Obviously there are others in a better position to judge than me about that. By the way, save states would be nice for TG16, megadrive, snes and a few others too, just not as critical as the NES since NVRAM was cheaper and in greater use by the 16-bit era.

Don’t know how I feel about rewind support OTOH, seems a bit TOO flashy for MAME/MESS, but who knows. It might be useful if it could assist in what little driver development is left.

Quite correct that most of the NES stuff has already been done, but man would it be nice to see MESS/UME at the NEStopia level – FAR nicer, more convenient and “future-proof” than anything available today. Thanks to eta, we’re definitely a lot further along than we ever have been before.

I somehow seem to remember it being something along the lines of *wi$ or something like that. Does anyone know if this rom is emulated in MESS or anything else for the BBC micro?

I miss the gold days of MAME/MESS with weekly versions of a new u and significative progresses…
Today, more and more people is giving the soulder to the projects and unfortunatelly, is missing new contributors :/
Anyway, that made any simple progress very significative for us!

@proc: save states for NES in current version are way more stable than in u3 (previously gfx banks were not properly restored, while now they are, and all the mapper-specific variables are saved and restored properly) and I’m going to commit soon some code which improves their stability even more.
however, some specific games still refuse to reload properly (maybe something not saved in the CPU?) and I haven’t been able to find the problem, despite 2 weeks of attempts…

save states are high in my priority list (I suck in a lot of game genres, so it is a pain sometimes to test in search of bugs), but since i haven’t ironed out all the problems I did not want to mention in the whatsnew something that is not fully usable

@Haze: I also think that OG fixed some bad bug in the .dsk format support, so probably PCW and CPC disks would work better than in u3…

Hi Haze,

Concerning olds103t, it is not a bootleg but an “official” version made by IGS for Tencent games and has licensed for some PGM games like Martial Masters.

Appreciate the work done on the BBC cassette interface, I can finally use the softlists I submitted many moons ago. Hope to soon see the same for the FDC to encourage me to submit a floppy softlist :)

shideravan> the problem is if you want sustained weekly progress then you need a whole team effort to do it.

the work from a few weeks ago with Cool Riders, Gunpey, Popobear etc. shows that there is still room for such levels of progress to be possible, and to a point the flurry of Playstation work shows the same it’s just you need more developers with that level of drive if you’re to give the impression those levels of work are being maintained while others are taking a break to reevaluate things.

there is plenty to do, a lot of it might not be interesting work, a fair amount of the work also involves huge undertakings (and people shouldn’t be under any illusion, if you make changes to certain areas of MAME/MESS then you have a HUGE responsibility because so many things depend on certain CPU cores, devices or the MAME architecture etc. which is no bad thing but can make the scale of the work intimidating if you’re not up for it)

For all I’ve said about it being important to have experts for certain systems above, it’s also important to have developers who will tackle anything, even if they don’t care about it much. One of the issues I’ve run into with the Fruit Machines is that nobody really wants to help write actual code and help develop / debug / trace driver / CPU problems simply because they don’t like the games, and that’s a shame because the technical challenges are just as interesting and a team effort would still go a lot further towards getting things done than everything falling on my shoulders.

Even if I look at something like Raiden 2, a popular driver / title I still get the impression that Mamedev have in the past figured out far more complex things, but emulation is taken so much for granted these days that everybody is just of the belief that somebody else will eventually do it ;-)

Of course you shouldn’t underestimate some of the work that has been done, but mostly gone under the radar due to being on systems few people care about, or not yielding immediate results. The x86 stuff being done in MESS is going to be absolutely vital for so many things it’s not even funny, even if right now people are just using it to see what Operating Systems they can coax into booting ;-)

That is always going to be a problem tho, like I’ve said, I’ve struggled to find people to work on the fruit machines with, and even if we make progress it WILL go under the radar for a lot of people, even if it might actually be more substantial progress than Cool Riders ever was. Likewise things like the M20 work are probably phenomenal to see for people who cared greatly about that system, but when so many are wondering when MAME will handle Model 2 properly, or MESS will run N64 at speeds worth a damn then it tends to be overlooked by larger groups which is a bit unfair.

I do try to promote all the work here, although it’s not always easy because for some of it not even I have a clue as to the real benefits, or nature of some of the systems involved.

If anything maybe I should say the Cool Riders progress came thick and fast because once the initial difficulty was figured out the rest wasn’t actually THAT difficult, although I also get the distinct impression that had anybody else studied it in any depth they would have concluded the same thing which kinda indicates a lot of people aren’t even looking.

All in all, I guess I just don’t know, on one hand a lot of great work is being done, on the other, a lot of potentially great work is being left not done. The fact that we can still put out monthly releases and have some talking points is probably still a good thing, you’ve only got to look at some other projects and see they can go an entire year, if not more, without contributions. Something like MAME (especially with MESS) is at least open ended enough that there will always be some buzz around it, something to improve / contribute even if it’s not always the things people want to see improved.

Of course with a little more effort and some fresh devs things could really be flying, but the climate we’re in now suggests as much as anything people no longer have time, and the fresh devs with the required understanding / knowledge simply don’t exist. You have to remember that emulation is something of a dead-end specialist task, an unrecognized contribution to the world with few genuine job prospects especially if you actually care about doing things in a proper open way for everybody to benefit rather than trying to sell your work to the highest bidder. It’s hardly surprising therefore that it’s not an avenue that many people explore, especially if they have no affinity to or knowledge of anything older than a Windows PC / Mobile phone in the first place.

Thanks for your insight on NES save states, eta. It’s good news indeed. Have you been able to remove the message, “Warning: save states not officially supported for this game (should be ‘system’)”? Either way, that’s going to make a big difference in usability and should hopefully lead to a lot more people using the NES driver over others, especially when UME .149 hits. I hope it also helps you with the PPU rewrite, whenever you have time to tackle it. :-)

Haze, I was just wondering, what’s your view on the things over at nonmame?

could you be more specific, is there something I should be aware of?

I can be more specific, but I was trying to avoid it :)
I have been following the releases of MAME for many years now, but haven’t been following the development behind the scenes for so long. I only just recently had a reason to do so.

From many of the discussions I get the impression that the pace is slowing down and there is not a whole lot left to work on that hasn’t already been done.

From a user point of view what would be an improvement at this point would be not having to download and keep track of so many different emulators.

From what I can tell, some of the emulators on NonMAME are Open Source and others have been around for many years. I realize the listed systems are there for a reason and must be the hard cases and that the listed emulators probably have taken shortcuts MAME cannot take. Still it would be interesting to hear your opinion on the current status of these systems and what can and cannot be done for MAME.

Well.. let’s go over them..


MAME does laserdisc games, but only a few have the capability to dump to MAME standards, plus I believe some people are working on better standards, so it probably doesn’t make sense to create 20GB CHDs for each game at this point.

Legal mumbo-jumbo aside I’m pretty sure any driver for any of the games would be accepted, the Dragons Lair (non-working) driver has been in MAME for as long as I can remember, and even a perfect rip of a laserdisc falls well beyond what the current commercial offerings of the game give you.

From a dev perspective these things are even less interesting to work on than the fruit machines however, even the game code is uninteresting, not to speak of the games. Most of them just string together video clips with timed inputs, I’d consider them absolute bottom tier arcade games myself. Combine a lack of dumps of the quality required for accurate emulation (which the Daphne MPEGs certainly aren’t) a lack of a true standard for creating such dumps, a lack of people to do them, certain devs trolling about legal issues, and completely uninteresting hardware and you have your answer ;-)


A lot of that is in MAME, but without an SH4 recompiler it’s a non-starter in terms of performance, and I don’t think anybody is going to work on compatibility while performance is so bad in MAME. AFAIK they haven’t shared all their Naomi 2 knowledge either and Demul isn’t open source. Of course a fair amount of the understanding of the encryption allowing these to work at all was thanks to the work of Andreas who is part of the MAME team ;-)

DICE / Laser

MAME has the architecture for doing some simple discrete stuff and as you can see, runs pong to some degree these days. The other titles are more complex, and working on this kind of thing requires a special level of understanding, it’s nothing like traditional emulation.

Dolphin / Dolphin Triforce

Nobody has put any effort into emulating the GameCube in MESS, which would be a precursor to any of this. Even if they did I doubt performance would be up to much (although there is at least a PPC recompiler) I doubt you’re going to see anybody give this priority, plus I believe most of the Triforce ROM based dumps are just ugly hacked up ROMs.

Final Burn-FF

Don’t really know what that’s on their list, I wouldn’t consider force feedback to be worthy of an entry. MAME doesn’t have it because there is no generic, non-hacky way of doing it, it’s more of a ‘mechanical’ thing than any of the fruit machines in most cases. It’s possible MAME could expose the outputs for external hookup (maybe it even does already) but I guess that’s not what people want.


obviously most of this doesn’t fit, but the site lists Monaco GP. The ’emulation’ there is just a recreation of the game, for actual emulation see DICE / Laser entry.

Radikal Bikers Emulator

does anybody still use this? performance is fine in MAME if you have a decent machine.

SEGA Model 2 Emulator

closed source, complex, a few devs do have the source, but ElSemi’s work isn’t really portable (apparently lots of inline x86 stuff, caching interpretor cores, not really compatible with MAME) I think most devs are waiting on those with access to the source to do it, and those with access to the source aren’t doing it.. I wouldn’t trust the existing MAME CPU cores here, because ElSemi fixed many bugs since the ones he did contribute.


MAME already runs Scud Race, but slowly (about 40% on my C2D ingame 32-bit build, if you wait through the painfully slow startup which ironically is *apparently* caused by the way the multithreaded poly code caches things as part of a speed optimization) It could probably do with that Software 3D Renderer for GPUs that people talk about… I’m sure some fixes could be ported across for other games if needs be so that everything boots mind you. Some games have encryption on the gfx (not working properly in SuperModel) seems to be the same scheme Naomi uses, which is also used by some ST-V games it seems. If there is progress on that front it’s more likely MAME will be helping Supermodel than Supermodel helping MAME.

Obviously the SuperModel code can’t be used directly for any kind of 3D acceleration, because it’s more traditional 3D acceleration and doesn’t fit the model MAME would require (effectively programming the GPU to be an accelerated software renderer)


Again MAME is fine on a high end system, and I’d struggle to even call VN an emulator, game physics are all wrong because they HLE half the main game code for speed.


This stuff has been slowly improving in MAME, although still lacks collisions in Malzak. Afaik the emulator is just a big hack full of offensive comments anyway ;-)

nothing too special, interesting or easy on there AFAIK.

Big thanks for the write-up! It gave me some more understanding to the situation here.

A few notes to add, Haze:

* Quite right about Demul.. it’s highly unfortunate that it’s closed-source and restricted to a single platform – so much potential wasted from such a historically significant hardware series. I get questions about that sometimes (as if I have the answers!). We remain hopeful the devs will change the situation but not holding our breath.

* Discrete circuitry emulation requires a certain skillset that’s even more specialized than general emulation demands, so unless the likes of Couriersud and Derrick return to the fray, the situation in MAME is unlikely to change anytime soon.. it’s not exactly attracting hoards of devs on that front.

* Final Burn-ff is included because force feedback was so critical to the game experience, especially for games like Outrun (think fast – what things do you remember most from playing it in the arcade?) MAMEHooker used to support FF hookup for Outrun but it regressed, and wouldn’t really be an optimal solution if it hadn’t anyway.

* Monaco GP is indeed primarily a simulation, but it is included because it’s widely regarded as a classic up there with games like Turbo (and I enthusiastically back that up!). Leaving it out would be a huge omission – it’s already mostly forgotten these days, and emulating such a sophisticated nest of analog circuitry accurately at playable speeds would be about as possible as emulating a PS3 on a Gameboy. ;-)

* Radikal Bikers and Rave Racer are too slow to be enjoyably playable on the current baseline test system (see site rule #2). The system is a mobile Core i7 Sandy Bridge with a 2.8ghz CPU, which judging from those who have written in with their specs recently, is about on par with that of the average reader.

* WinArcadia/Signetics targets are probably the easiest on the list to hit ATM… not sure why Jacobs won’t help MAMEdev as he seems to have a lot of the issues sorted out and has declared his intentions to help.. but w/e. Regardless, not likely to be an issue for too much longer anyway.

re: outrun, I really don’t remember it having Force Feedback at all if I’m honest with you, definitely not a memorable aspect of the game if the units I played had it at all. What I remember of that game is more the general feeling it created, the graphics, the music and a sense of a cool breeze running through your hair, just like being on a beach, it certainly wasn’t anything like Afterburner where you simply can’t forget being tossed around.

For Radikal bikers I’d be surprsied if there isn’t a MAME build capable of running at full speed on such a setup, even if maybe it’s an older one. For Rave Racer again there should definitely be builds capable of running it better than VN, I simply don’t count VN, it isn’t so much an emulation as a (bad) port.

re: WinArcadia, I believe our last contact with the guy was him accusing of of ‘stealing his code’ when we emulated the same things as him at one point (he seems to consider the roms he embeds in his emulator as his properly, which is truly bizzare) I don’t think he has anything of offer that can help tbh.

Ok, chalk it up to a matter of opinion then. Force feedback played a big role in all the Sega games of that era IMHO – Outrun, Afterburner, and Space Harrier being the most prominent. I recall the FF in Outrun being the best of all, it was certainly the first driving game I had ever played which had such a feature, and in the sit-down cabinet – wow.

I haven’t come across any MAME builds including older official ones that could play Radikal Bikers or Rave Racer at an enjoyable speed. Not saying it doesn’t exist, but I haven’t seen any. I agree with you about VN though – it’s far from an optimal or accurate experience but better overall than MAME ATM if you just want to play it at a consistently smooth framerate with similar specs.

Jacobs once said that he had regretted taking the position he did towards the dev team, and that he is indeed interested in helping – but that was over a year ago and I’m not holding my breath at this point. You’d have to contact him for more information if you were curious. I agree that he probably doesn’t have much to offer though, especially now that the Signetics driver has significantly improved over the past year. On the MESS side, however…

Hi, dear Haze
We managed to dump roms PCB game TIME PILOT.

We would like to put in MAME

Indian Battle of TAITO OF BRAZIL was also dump for me

The following files and photos PCB

Roms files and photos pcb already in the hands of the hap and smithdogg

thank you

Yeah, it will be added to MAME by “the hap”(:P), looks like another odd conversion to Galaxian hardware.

The color prom still needs to be dumped

aw, the hands of the hap were not good enough to make the game working nice in MAME today

I thought Time Pilot was a (distant) relative of Galaxian anyway?

hmm, no, I was thinking of another game ;-)

Hi Haze,
Sorry to be totally off topic with the following question. I’d have contacted you via an email address, but I’ve not managed to find one for you. At any rate, I’m a freelance journalist for the UK print magazine Retro Gamer. I was given Corrado Tomaselli’s name by Smitdogg of the Dumping Union during research I’m doing for an article, but I’m struggling to get in touch with Corrado and thought you might be able to pass a message on to him.
I’m trying to verify that Chris Stamper developed Vortex, and Smitdogg thinks he remembers Corrado telling him that this might be the case. I was thinking Corrado might have found Chris’s name in an unencrypted version of the game’s source code, and I’m keen to know if this is the case and also if the code stored the names of any co-authors of the game, or messages from the developers.
Any help you could give me would be gratefully recieved, many thanks for your time,
Rory Milne.

Vortex is by Zilec, but there are no obvious names stored in plaintext anywhere in memory space.

I’m not sure what encoding the rest of the text in the ROMs uses, so I can’t at this point check if there are any hidden strings.

I think the assumption just comes from it being a Zilec game, and there being known ties between Zilec and Chris, rather than any solid evidence of who developed this specific game. It’s a very obscure board (previously thought to be a Taito game, but in fact is just a conversion of a Taito PCB) so I doubt you’ll find too many details about it, if any.

I doubt Corrado would have any additional information either I’m afraid, he typically just dumps things, he isn’t a programmer / developer, and there is nothing obvious to find in the ROMs.

There is probably a colour version of it out there somewhere too for conversion of the colour Space Invaders boards, assuming there is no way to get the current set to output data needed for a colour display (there doesn’t seem to be) We’ve seen screenshots suggesting as much at least.

Hi Haze,
Thanks for getting back to me so quickly and with so much interesting information. It’s a shame the Vortex developers can’t be identified as yet, maybe one day, though.
I’ve seen the colour Vortex screenshots too, and like you, hope the existing dump can be made to display in colour at some point or that an alternate dump manages the the feat, thanks again, Rory.

@proc: with latest changes to svn, NES and GB/GBC drivers should support saving & restore quite well
also, after several tests, I never had issues with SNES and Megadrive either

so if you can find any game that has still issues with save states for any of the systems above, please let me know

I’d love to fix the remaining problems…

Wow, that’s really GREAT news. Thanks, eta! Hopefully you were able to remove the “Warning: save states not officially supported for this [system]” message. I noticed it appears even for MESS drivers that seem to work fine with save states, like SNES and Megadrive, as you said. I had previously observed big issues with the NES and TG16 drivers, and since SNES and Megadrive throws that message too, I figured those were equally borked. Perhaps not – but removing the message for drivers that work would avoid confusion I think.

Can’t wait for .149 to check this stuff out!

if there are no reports of games mis-behaving when restored, I’ll be more than happy to make the message disappear

but given that there are thousand of games and I only tested a dozen for each system, I believe we need some more testing, before removal

hopefully, we will get enough feedback before next release so to remove the warning message :)

oh, and in case you are not aware of this, you can test the changes too by using BobZ’s daily build


let me know if you find games with problems

I can’t test this build right now, but I’d like to go ahead and say that I for one am very interested and thrilled with the current state of Playstation emulation.

Ok, I’m back with my useless evaluation.

CDDA on PSX is still broke (Ridge Racer, etc) but it doesn’t produce horrid noises, so that’s good. Games which previously didn’t boot now do (Incredible Hulk, Golden Nugget). Nothing else of note…no GPU improvements or anything.

In other random notes, FDS now works, FM Towns is broke, and the Megadrive hashlist is becoming mind boggling to keep up with, so I hope that the format won’t change drastically anytime soon.

PSX work is ongoing, were there specific GPU bugs you were hoping would be fixed? (a lot HAVE been fixed but if you were looking at something specific it would be best to mention it)

CDDA, people have mixed results with, we’re pretty sure that the MAME CD .cue/.bin reading code parses certain images incorrectly right now so can create bad CHDs in such cases and also might throw off some drivers including the PSX

If FM Towns has regressed in a significant way I imagine Barry will be on it, it’s not a system I know enough about to test.

Madden 2002 does a number of strange things that I assume are GPU speed up tricks. I really can’t describe the glitches sufficiently, but the game is playable.

I didn’t really expect anything new to happen, but I wanted to mention it in context of the driver showing very noticeable improvements recently.

> Megadrive hashlist is becoming mind boggling to keep up with, so I hope that the format won’t change drastically anytime soon.

care to explain a bit more what you mean?

I’m not sure how to reproduce most of the new ROM changes. I understand the goal of PCB accuracy, but I’m hopeful that my customized 148u3 hashfile will last a while.

Re. WinArcadia, the source code is freely available and there are also various documents at my site explaining the memory maps, etc. I am too busy with the various Amigan Software projects to work on MAME/MESS/UME as well, that doesn’t mean someone else couldn’t do it however. I don’t intend to abandon Ami/WinArcadia for various reasons, but I’m happy for all the various emulators of Signetics-based machines to continue to improve.

@Tim: sorry for the late reply. anyway, in most cases you just have to split the images with the new sizes

if it was a 3MB set and now you have a 2MB and a 1MB roms, then it’s just a matter of opening the file in an hex editor and separate the file in the two chunks

in one case, a 3MB rom was enlarged to 4MB because the chip was verified having that size. if you check the corresponding xml, you will see a comment saying that the rom is padded with 0xff. this means that you have to create a new file with an hex editor containing 1MB of 0xff bytes and paste it at the end of the original rom so to pass from 3MB to 4MB

this would bring you to having the new set except for 1-2 roms whose splitting is a bit more complicate (typically, splitting the roms in half and then deinterleave the files so to split even and odd bytes)

I’ll be happy to give you further assistance if needed. just drop me a PM on one of the forums I visit (MW or bannister)

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.