@Sun_Crypt (Liquid) Kids! Do you enjoy those long posts about the Sega Saturn in which I tackle a weird subject matter at length thanks to a well-researched explanation? Well, you’ve come to the wrong place today, because here’s blurry corporate shenanigans about which I have more questions and personal hypotheses than proper answers instead.
12/30 I had to go back and edit the last four posts because I forgot the birthday cake emoji (also: that weird partnership business mess between Sega and SNK)
Here’s a bit of Saturn’s history about which I am bummed I don’t know more. I know people who work at Sega and at SNK, but I’ve never thought of solving this old crossover problem by using my connections in either of the two camps. I’ll have to see if I know anyone knows anyone who knows anyone who might be able to point me in the right direction. But here’s what I know for now.
In September 1995, probably around the time of the Autumn’s AOU Show, an event for the arcade gaming industry, Sega and SNK announced an exceptional “cross-license” partnership giving SNK the right to publish games on Saturn and Sega the right to publish games on NEO•GEO.
The way the deal is presented on these pages, from 1996 onwards, SNK would publish its own ports of NEO•GEO games exclusively on Saturn, under the pretext that this was the only console they were confident of getting a flawless port for their games.
Of course, there had already been numerous ports of NEO•GEO games to Mega Drive, Super Famicom, Game Boy, Game Gear, X68000 etc. licensed by other publishers, first and foremost Takara. But SNK, for its part, had stopped publishing on other manufacturers’ consoles as soon as the NEO•GEO ecosystem gained traction, and the company had just publicly reiterated a year earlier, at the launch of the NEO•GEO CD in September 1994, that they themselves would not be releasing games on Saturn and PlayStation. So this was quite a coup for Sega.
At the time, no one was fooled: this partnership took place mainly because the NEO•GEO CD was a flop, and people were rightfully expecting more SNK support on Sega’s hardware than the other way around. But officially, Sega and SNK would be working hand in hand to support each other’s machines. In the interview above, the SNK representative explains that Sega is a long-standing arcade partner of SNK (by which he means that SNK games can be found in Sega arcades), which greatly facilitated the entente cordiale between the two companies.
I remembered this interview quite well, and more generally the treatment of this news in Sega-affiliated media, but until now I’d never bothered to check how SNK’s local Pravda, Neo Geo Freak magazine, had covered the news. All in all, they only mentioned it twice: in the Neo Geo Freak issue released in October 1995, which reported the news rather soberly (below)…
… Then, in the end of the year issue, through an article summarizing “the ten biggest NEO•GEO news stories of 1995 as voted by our readers”. it’s the news of the partnership with Sega that comes first, with a brief comment “well, yep, indeed, that’s a big one” before awkwardly skipping to a less sensitive subject. You can read between the lines in both articles that this announcement was a rather delicate point of contention with the magazine’s hardcore fanbase.
Odd thing #1 :
Here is the result of this partnership:
Games published by SNK on Saturn → fifteen, not counting reissues and compilations.
Games published by Sega on NEO•GEO → zero.
So, did Sega really give no guarantees in this partnership deal? Not even a little bogus game on MVS to give the change, like a port of Columns, or Tanto-R Goes Hawaiian ? Are there any prototypes of Sega games running on NEO•GEO hardware in some drawers ?
Odd thing #2 :
You can see on the right-hand page of the first article above that the first two SNK games announced and released on Saturn would be Fatal Fury 3 (arcade release: March 1995) and KOF’95 (arcade release: July 1995). Just to clarify, the other titles shown below these two are just to give a glimpse of the NEO•GEO catalog. It’s speculated at the bottom that “the third episode of Samurai Spirits” (still unannounced at the time of going to press) might be the third game, but it’ll turn out to be World Heroes Perfect after all.
KOF’95 was released first, in March 1996, with a dedicated ROM cartridge and a wonderfully polished port. The game was available on Saturn even before KOF’96 was shown on NEO•GEO, so it was still fresh news. It’s a pretty flawless version of KOF’95, even if Sega and SNK are a bit optimistic when they boast about the port’s 1:1 fidelity.
Fatal Fury 3 was released later, in June 1996, without a ROM cartridge, via a port that we’ll politely describe as fetid, while the sequel Real Bout had already been released in arcade. However, of the two games, it was Fatal Fury 3 that ended up as the only true Saturn exclusive, because…
Odd thing #3 :
… The “exclusive” and privileged partnership between Sega and SNK won’t even last a year. Just after the release of KOF’95 in March, SNK announced its debut on PlayStation with a port of KOF’95, and no cartridge needed thank you very much. The game was released on PlayStation in June 1996, at the same time as FF3 on Saturn. It’s obviously an inferior port to both the NEO•GEO CD and Saturn versions, but it’s genuinely good otherwise.
(And now is a good time to share this excellent comparison video.)
Thereafter, SNK ports would usually be released on Saturn one to three months ahead of the PlayStation, but one gets the impression that this is mainly because it was quicker and easier to port games to Saturn, or because SNK had the right staff available for the right game, and not because of any strategic preference or promise made to Sega.
A handful of games were even released on PlayStation ahead of schedule - notably SamSpi, which makes me wonder if this tidbit of SNK’s business has anything to do with the staff involved in SamSpi eventually leaving to develop Guilty Gear on PS1 at Arc System Works. But that’s a discussion for another thread.
Odd thing #4 :
The release of KOF’95 in the West. It was SCEE themselves who were to publish all SNK games in Europe in 1996, instigating the rumor that a regional exclusivity deal existed between Sony and SNK, and that SNK games were blocked on Saturn in the West. It is true that Real Bout remained a PS1 exclusive outside Japan, and KOF’95 was not released on Saturn in Europe until 1997. Neither game came out in the US on either platform, because your Sony sucked balls.
I gotta say, This Euro-exclusive exclusivity is not impossible, but it seems a bit incongruous to me. PlayStation was already pummeling the Saturn in Europe by deliberately downplaying the importance of 2D games; what need was there to join forces with SNK (of all people) to keep pressing on Sega’s neck?
I rather suspect that Sega of Europe was in a real financial bind by then, and that releasing a game requiring the production and ordering and storage of a dedicated cartridge was quite a mess, and that they (first) refused the distribution deal later promised by Sony Europe. But after all, Sony was all in on killing off the competition by any means necessary back then, so why not.
Odd thing #5 :
The developers of these first two games; they were clearly chosen by Sega.
The main developer of KOF’95 is Rutubo Games, the best arcade game porting studio at the time, especially when they originated from 68000-based systems. We owe them X-Men: Children of the Atom, After Burner II, Out Run and Space Harrier on the Saturn. They were very close to Sega at the time.
The developer of Fatal Fury 3 is Sims, another good Sega buddy, who frequently helped them on Mega Drive and Mega CD. It’s a studio with less talent(s), but also generally less time and fewer financial means to accomplish its ports.
Putting KOF’95 first isn’t surprising in itself: it was probably the biggest hit in SNK’s history, and was the port most in demand by Saturn gamers according to magazine polls, while FF3 was poorly received by gamers and overall a flop in the arcades (hence the express release of Real Bout less than a year later).
But it’s odd to have announced these two games together as the standard-bearers of this exceptional and exclusive partnership, only to squander the second game without even keeping up appearances, Japanese-style, and granting it its own dedicated Twin ROM cartridge. A technology that the PlayStation was physically incapable of offering.
Was it Sega who dropped the ball too quickly, on learning of KOF’95’s release on PlayStation later in the year, or in response to SNK’s deal with SCEE in Europe?
Was it SNK who cut the deal in the face of its own slowing business in 1996, Sega’s lack of motivation to release games on its system, or due to being annoyed by the growing competition from the ST-V in the arcade? Does this whole consumer business mess conceal a failed rapprochement in the arcade field between Sega and SNK? I don’t know. I don’t know! Rah! I don’t know…
Odd thing #6 :
KOF’95’s notorious Twin Advanced ROM System is a Mask ROM cartridge capable of loading part of the game’s data into memory. In the case of KOF’95, these are mostly graphic assets helping with sprite animations, stored on a 16 Megabit ROM (about 2MB).
This Twin ROM method was used for only one other Ultraman game, by Bandai.
SNK followed KOF’95 with two weak ports, FF3 then World Heroes Perfect a few weeks later, before Sega and SNK adopted a new universal RAM cartridge method allowing any game to use the extra RAM in its own way. Other publishers followed suit with games allowing or requiring the use of a RAM cartridge.
I think you already know this if you’re the type to visit the Insert Credit forums, but two versions of this “extended” (kakuchō) RAM exist: the Kakuchō RAM Cartridge (which adds 1MB of RAM) released in September 1996 for the port of Real Bout Fatal Fury, always around to fix its elder’s bullshit, and the Kakuchō RAM Cartridge 4MB (doing what it says on the tin) released in November 1997 for the port of X-Men Vs. Street Fighter.
The idea of boosting a console’s RAM through an external peripheral to improve the performance of a CD-ROM game first appeared* with the Arcade Card created by Hudson for its PC Engine ports of NEO•GEO games, starting with Garō Densetsu 2 (Fatal Fury 2 ) and Ryūko no Ken (Art of Fighting) in March 1994. This process has the obvious advantage of requiring only one card to be produced and sold for all subsequent games. Nintendo took up this idea for the Nintendo 64 with the Expansion Pak.
*The influence of early consumer personal computers such as the MSX, already adept with performance expansion cartridges, is debatable, but the direct filiation is less obvious.
For the past three decades, the conventional wisdom behind the abandonment of Twin ROM technology in favor of a universal RAM cartridge has been that it is a more economical solution for gamers and publishers alike. But, while re-reading the gaming press of the time the other day, I came across a rather enlightening interview with Sega’s hardware team in charge of the Twin ROM cartridge, which at the very least puts this simplistic answer into perspective.
First of all, throughout the article transpires the impression that Sega took care of absolutely everything regarding KOF’95 port. This is in stark contrast to the interviews for subsequent games, which usually involve only SNK staff.
Next, in the interview, Sega explains how they came to choose this Twin ROM system and, thankfully, the journalist has the intuition to ask why they didn’t choose a RAM solution like Hudson on the PC Engine instead.
Sega answers that, in truth, producing a universal RAM card would cost them extraordinarily more than a game-specific Mask ROM system would. Basically, a Twin ROM cart like the KOF’95 one uses the same production principle and components as a Mega Drive cartridge. Not only is it much cheaper in pure COG terms, but Sega already has the components, suppliers, production lines and so on. They even imply (unless I’m over-analyzing their answer) that it might be convenient for some departments at Sega to continue cartridge production lines via this kind of hybrid game.
And you know what I think? In my opinion, the real reason Sega dropped the Twin ROM was because it was far too complicated to develop games this way, and would have considerably slowed down third-party ports (and/or forced Sega to take care of everything, as with KOF’95).
We now know from more recent interviews that Sega had canceled at the last minute the possibility for a Saturn to read games from the cartridge port, for fear that Hong Kong pirates could exploit this vulnerability to run game copies (in 1994, burning and selling a pirate CD-ROM was not yet economically plausible).
When one digs deeper into how the Twin ROM communicates with the Saturn, one realizes that Sega and Rutubo are, in a way, self-hacking the Saturn by bypassing these protections in place, and making the console believe that the cartridge data comes from the CD, by (simplifying here) exploiting a little-known feature of one of the console’s too many processors.
This also explains why an amateur developer recently succeeded in creating an image of the KOF’95 CD-ROM including the entire contents of the Twin ROM cartridge in the game disc, and instead using a 4MB RAM cartridge to store the Twin ROM card data claimed by the code.
Now, when you further look at the esoteric path Rutubo took to synchronize the cartridge and the disc, it’s a mind-boggling mess, all the more so for running all this on a hardware with notoriously Escherian programming logic. Developing with the Twin ROM method is probably a crime against the Geneva Convention of Game Programmers.
I therefore posit this alternative explanation: Sega realized that the Twin ROM method was certainly economically viable from a production cost point of view, but untenable in the long term in the face of the abilities of third-party developers less familiar with the mysterious ways of the Saturn, compared to the clearer and more “brute force” solution of just adding RAM and go suit yourself. It’s also possible that Sega didn’t want to share so openly with everyone how they’d bypassed their theoretical protection with the Twin ROM.
One last thing (goddamnit, every time I pick a topic thinking « good, good, that one will be short and sweet »), here is a small detail that made me laugh as I was trying to find out if the Twin ROM mysteries of yore had since been solved: the first result Google gave me was an old discussion on the Segaxtreme forums trying to solve this mystery over twenty years ago, in January 2003, and I didn’t realize until the second time I checked the convo that the question had been asked by a young and naive “CHAZumaru”. Plus ça change…