Monday, November 10, 2014

I Wanted One Last Project to Give This Site a Proper Sendoff, So I Wrote an Original Soundtrack Concept Album for a Manitoba-Themed Nintendo Game That Doesn't Exist

Hello there, friends! I'd hate to bury the lede too far down under my characteristic exposition, and the post title gives you the gist of the idea, so let's get straight to the important part: I'd been secretly writing an album off-and-on over the past year or so, and now that album is out.

Manitoba Circuit OST / マニトバラリーレース大冒険!! こんなに平らでいいのか?! OST
Pay-What-You-Want |

So that's pretty great. That's a pretty good feeling! The album is free to stream, and also free to download, though you are welcome to pay money for it if you find you enjoy it.

As I mentioned, I'd been working on this project in secret for a while now -- because, to my chagrin, I have noticed that the more I talk about an idea the less likely I am to work on and finish it -- so now that it's done and released I'll let the dam burst on it, broken into sections, while we're all here together one last time.

What is Manitoba Circuit?

Manitoba Circuit was a completed but unreleased Nintendo Entertainment System video game, partially financed by Tourism Manitoba, that was developed from roughly September 1987 to February 1988 and originally intended for a July 1988 release.

Home video games had come a long way from the dramatic industry crash of 1983; the juggernaut Nintendo Entertainment System was selling millions of units each year, to the point that one-third of North American households would own an NES by the end of the decade.

Advertisers, unsurprisingly, took notice of the market potential. Not only did some NES games feature in-game advertising, but development companies created game tie-ins for major motion pictures, and ultimately moved into video games created solely as advertisement for brands and franchises like Walt Disney World, 7-Up, Dominos Pizza and McDonalds. (It must be noted, of course, that this was not entirely new territory; corporate-branded video games date as far back as the Atari 2600.) There was even almost a California Raisins game, cancelled in 1990 for the probably-obvious reasons of "California Raisins" and "1990".

Sensing a prime opportunity for increased worldwide exposure -- and particularly in Japan, a source of tourism revenue already proving invaluable to Prince Edward Island -- Tourism Manitoba, under then-Premier Howard Pawley and his New Democratic Party government, embarked on an ambitious pilot project to recruit a team of game programmers from Japan and release a Manitoba-branded Nintendo game across all three main worldwide markets (Japan, North America, and Europe).

The eventual seven-man development team, comprised mainly of recently-graduated students and less-established programming staff, received a very warm welcome by Tourism Manitoba upon their arrival; they were shown the Winnipeg Art Gallery, given a tour of Assiniboine Park Zoo, taken for a night at the Royal Winnipeg Ballet, and even gifted tickets to a Winnipeg Jets game. This was very exciting for all of the staff, but most of all for lead programmer Tsujimoto Kenichi; his father, Tsujimoto Taro, was a former hockey player who had been drafted 183rd overall by the Buffalo Sabres in 1974.

TOBA SOFT, as the all-Japanese team dubbed themselves, were inspired by the newly-established "driving game" genre -- which focused more on the enjoyment of the driving experience than on the breakneck speed and arcade action of the racing-game genre -- and particularly by Rad Racer, the NES answer to SEGA's OutRun, which was released one month prior to their initial development brainstorming sessions.

To get a better idea of the area they'd be representing in their game, the team took a rented minivan out along some of Manitoba's more prominent highways; being from a small and mountainous country, their collective surprise at the topography of the area would later influence their title for the game's intended Japanese release.

Manitoba Circuit was programmed over the next six months and all but ready by February of 1988, even receiving the Nintendo Seal of Quality; all that remained was the physical manufacturing of the cartridges and the subsequent marketing efforts to retail chains and to customers. TOBA SOFT and Tourism Manitoba pegged the physical side of the game's publication at an an estimated three to four months, not unusual for the era, which would have placed the game at an expected early-July 1988 release and just ahead of peak summer-tourism season.

No one, of course, expected what happened next. The NDP majority government fell unexpectedly in March of 1988, when a backbencher defection defeated the party's budget on a confidence vote and triggered a surprise snap election. Seven weeks later, the NDP were out of power entirely, the Manitoba Progressive Conservatives forming a minority government under newly-elected Premier Gary Filmon.

The incoming Progressive Conservative government, which had gained power on the promise of vast cutbacks to save revenue, spelled doom for the Manitoba Circuit project; the new government considered the expected production costs of the final release an unreasonable expenditure, for what they saw as a meagre opportunity in an unproven market, and quietly cut its intended funding -- effectively scrapping the project entirely -- in late May of 1988.

Disappointed and upset by this abrupt change of fortune, all seven members of the TOBA SOFT team returned to Japan shortly thereafter, never to return. And Manitoba Circuit never officially saw the light of day, though it is rumoured but unconfirmed that altered versions of the game made appearances in unlicenced multicart compilations.

Although the experiment was ultimately doomed by the unpredictability of the era, the Manitoba Circuit project remains an interesting case study in alternative promotional activi--

Okay, No, Hold Up. Did You... Did You Make All That Up Just Now?

D'ahh, heck, y'got me. Pure fiction! You saw right through it, huh? I should've known nobody would believe that the NDP would try to impress people with free Jets tickets.

Anyway, no, Manitoba Circuit wasn't a real thing. But I did spend a year or so writing a soundtrack for it, and now we all have a story for it, so it's real in our hearts, y'know?

Really? A Year or So?

Well... maybe not exactly. That's more an estimate.

No one ever ends up exactly where they expected -- he writes, from a lakeside cabin near La Ronge, Saskatchewan -- and this is true of this album as well; my Manitoba-themed Nintendo Entertainment System racing-game soundtrack was, in its original inspiration, a Winnipeg-themed Super Nintendo Entertainment System role-playing-game soundtrack.

The more faithful readers of this blog may well remember where I was, life-wise, a little over a year ago; both of my two ongoing library employment positions had been unceremoniously eliminated, I was fresh off of a particularly crushing relationship break-up, my making rent on my apartment from month to month seemed increasingly unlikely, my local job prospects were looking gruesome and grim, and my mother was battling cancer. So it... wasn't a high mark, let's frame it that way. What better time to escape into music, then, right? Had to fill the days somehow, after all. ("Life is a shipwreck," Voltaire once wrote, "but we must not forget to sing in the lifeboats.")

Living in an alleyway apartment between Home and Arlington at the time -- what an amazing run of luck, in retrospect, that my car was only graffitied and smashed once that year! -- I had plenty of time to think about the abstract nature of neighbourhood character and place; I was located on the fringes of Wolseley, West Broadway, and the West End, and I was located within walking distance of Tuxedo, St. James, and the downtown core.

(Never underestimate my walking distance, incidentally. Especially on flat land. Which... describes just about everything in Winnipeg.)

All six of the neighbourhoods that I've just mentioned are very, very geographically close to each other, yet they all have very different characters and leave very different impressions, don't they? If you wanted to write retro-chiptune pieces for each place -- as I decided I did, once I learned that such a thing was indeed theoretically possible -- then, despite their being only a short bridge apart, you would choose very different musical styles for walking along a Wolseley Avenue and walking along a Wellington Crescent.

So I had the intention of writing a full Super Nintendo RPG soundtrack for and about Winnipeg, a collection of 100% system-authentic pieces for all of its major neighbourhoods and roadways and landmarks. I had a whole bunch of musical subjects picked out, and I had a grand unifying JRPG storyline for all of them (Super Winnipeg Saga: Invasion of the Purple City!!), and it was all very ambitious and exciting before two very important hurdles derailed the plan.

Firstly, I don't know how familiar you are with the JRPG soundtracks of the era, but at first blush you might not properly appreciate the sheer scale of them. Final Fantasy III/VIj, unquestionably one of the high marks of the genre, had sixty-one original compositions written for it; Chrono Trigger, another category-killer, had sixty-four. And though you wouldn't know it to look at the official soundtrack, which condensed multiple pieces into medleys, EarthBound -- unquestionably my favourite game of the era, and easily even now one of my top-five video games of all time -- had somewhere in the hundreds.

This would not have been as dramatic an issue if it were not for hurdle number two, which exponentially multiplied point number one. Hurdle number two is that -- and I do hope you will bear with me here, I've spent a lot of time researching these sorts of things since then -- the Super Nintendo sound chip, much to my surprise, does not actually contain any sound.

To clarify that statement: unlike almost all console hardware to that point, which could be programmed to output a default slate of notes and noises right from the start, the Super Nintendo operates entirely and exclusively in samples. Each and every SNES game has a special -- and, as per the tight Nintendo corporate controls of the era, pre-approved -- sample bank that the audio processing unit (APU) draws from, but without that individualized cartridge input, the SNES APU is functionally blank.

If you've ever wondered, this is why SNES games made by any given company sound like the other SNES games by that company, but sound completely different from the SNES games by other companies. (Compare a slate of games made by Capcom to a slate of games made by Konami, for example.) This is also why SNES games often sounded so peculiarly different from other SNES games, whereas games for the NES or for the Genesis always sounded like they were in the same sonic ballpark as other games on that system. I ended up learning a lot about the SNES APU, though unfortunately not a way to put it into practice.

Here, then, was my problem: my computers are gettin' up there in age, and writing with a special individualized sample loaded for every instrument is vastly more resource-heavy than writing for a single consistent soundset. (I know this because I tried anyway and the whole shebang, without fail, crashed on me in protest.) As well, certain SNES sample banks seem wholly disinterested in coexisting nicely with other SNES sample banks; as well (as well as well), all of the programs that can actually get these banks to play along with each other are optimized for newer Windows desktops, and the newest tech I have at my disposal is a three-year-old Mac laptop.

So if I wanted the grand SNES JRPG opera of my dreams, I'd have to guess-write sixty-odd musical pieces on hardware that couldn't play them, and then to play them I'd need a better, more powerful, proper desktop computer -- which of course was outside my price point, my price point at the time being "zero, because I'm unemployed". A dream for another time, perhaps.

However! When chiptune shenanigans close a door, they open a window, as I'm sure everyone agrees the saying goes. I revised my expectations, and what was originally a side-project idea became my new main-project jam, and do you know what happened next?

An NES Racing Game?

AN NES RACING GAME. Or rather, not a racing game, but a driving game -- and yes, as I alluded to above, there's a difference. I've done a lot of digital driving in my time, and I do a whole lot of highway driving in real life, so perhaps these sorts of things all work out the way they're supposed to.

Many of you will be familiar with OutRun, the glorious 1986 SEGA arcade experience that revolutionized the car-game genre; many of you will also be familiar with Rad Racer, the NES answer to OutRun, itself a major success in its own right. We never owned either of those when I was a kid, but somehow -- and I know we didn't buy it, so I'm thinking it was part of a pile that a family friend's son was no longer interested in, but I can't be 100% certain -- somehow, we did own Rad Racer II. Rad Racer II was not a success at the level of its predecessor, and I'm not sure it was even a success at all; in those rare times when it is thought of at all these days, it is thought of as a poor gameplay successor and a disappointing sequel. What it did have, however, is a soundtrack (of two in-game songs and one ending song) by a young Nobuo Uematsu, who would soon become one of the best-respected founding fathers of video game soundtracks as an evolving artistic medium.

I am most certainly not at the level of a Nobuo Uematsu, not even baby-Uematsu; I am also most certainly not at the level of a Keiichi Suzuki or a Hirokazu "Hip" Tanaka, the two Mother-series musical masterminds coaxing pitch-perfect genre-study pieces out of fascinatingly primitive machinery. But then, of course, they are exceptional; the vast majority of folk tasked with writing NES music weren't at that same level either, and they didn't let that stop them. A fair number of people composing music for the NES weren't 'composers', per se, so much as programmers coincidentally tasked with making sure the game had music -- and a fair number of NES composers wrote the music in pure hexcode, which is both tremendously impressive and wildly intimidating.

I mention all of this because my aim was to write an NES-era game soundtrack that would seem authentic, rather than simply 'plausible'. The soundchip has plenty of neat tricks and fun treats that weren't discovered until much later in the lifespan of the console, and that even then only turned up in maybe two or three games; this is because, over the decade or so that companies supported the NES, there were only ever a handful of people who A) had a solid grasp of musical theory and composition, B) developed a thorough familiarity with the capabilities of the sound hardware, AND C) had enough time in the development process to write and revise something nice. So even though I had the modern luxury of leisurely experimentation, I left a lot of segments on the proverbial cutting-room floor because they wouldn't've existed at the time; the chip could play them, but the people working the chip wouldn't have. You get me, yeah? Better to aim for the flawed but ambitious output of the era than to forge something meticulous but obviously out-of-place.

So Why a Manitoba-Themed Game Soundtrack?

Because I thought it'd be a really cool thing to have, and nobody else has ever done it. (You'd be amazed at how much of my output on this blog over the years has traced back to this exact reasoning.)

The reason most of all that I wanted to do this sort of project, something specifically musical, is because it was out of my element and I wanted to see if I could. That's part of why it took me as long as it did; it was a major challenge, and I had to learn a bunch of new stuff to do it. Which is good! It's important to challenge oneself, I think.

Having established what I've been working on, here's what I've been working in:

Deflemask is a cross-platform, multisystem hardware-accurate tracker; the GUI (graphical user interface) is obviously slicker and sleeker than the environments available at the time, but as I lack access to the original NES hardware kits and software compilers, I made the executive decision to shrug and roll with it. The end result is that all of these pieces are 100% authentically NES-compatible, to the point that they can play on an original NES if exported and compiled as such; I can even export the album to .NSF, if you're familiar with that format, but I still have yet to work up the nerve to do so. (It's a little galling to think of months of work condensing down to maybe 20kb of output. I don't know how you programmers out there manage it.)

Not only did I have to learn how to control the program and how to write tracker sequences -- it's like trying to perforate and synchronize four player-pianos at once -- but if you look along the right side of that screenshot, you can see how almost every note you hear on this album was customized and hand-drawn. I don't know if you've ever had the experience of sitting down and diagramming your waveform on the fly to get the sound you want; it's kind of a trip! It's fun to play around with it just to see what happens to it. It's also a lot of fun to chart out the folio options of the noise channel, piecing together thunderclaps and fireworks and other dramatic punctuations.

Speaking of channels -- and getting back to the authenticity aspect for a second -- you'll note that the tracker has five available channels, just as (or, rather, because) the NES soundchip does. However, of those five channels, the vast majority of NES music only uses four; the fifth channel is a rudimentary sampler, for very tiny samples, and is (almost) exclusively reserved for in-game sound effects whenever the gameplay demands them. (This channel is also responsible for any voice samples, though those tend to stick to the title screen.) There are certain exceptions to this -- Sunsoft games like Journey to Silius used a rad bass sample, and Super Mario Bros. 3 dedicated some cartridge space to drum samples -- but, again, such advancements are both relatively rare and considerably later in the lifespan of the console.

For the bulk of the NES library, then, and particularly so in the era I aimed for, the music is limited to four channels: two square-wave channels (generally used for leads and/or harmonies), one triangle-wave channel (almost always the bass), and one noise channel (for percussion, and occasionally for those atmospheric effects I'd mentioned earlier). I could go into a very long side discussion from here about how the challenges of limitations help to spur and refine creativity, but honestly it'd just be me quoting Chuck Jones' interviews on the subject for like five pages, and really I think this post is going to be long enough already.

So What's All This PAL Stuff on the Album?

Ha! Oh, man, PAL. PAL is hilarious.

This is its own special subsection of gaming history, so let me break into another quick technical breakdown, if I may. North American NESes and Japanese Famicoms run on a 60Hz internal processor; this is because both regions use (well, more accurately in our modern times, used) NTSC analog televisions, which run at a 60Hz refresh rate. European analog televisions, however, ran at a PAL 50Hz refresh rate, because the infamously-different electrical grids of the region only allowed for 50Hz power supplies. And because European TVs ran at a 50Hz refresh rate, European Nintendos likewise ran on a 50Hz internal processor -- which was now, unfortunately, tasked with running games and music designed for a 60Hz processor.

Effectively, unless a game developer made a specific point of fixing their programming for the PAL European release -- and an insultingly precious few of them did -- any game released for the European NES ran and played 17% slower than it did on the North American NES or the Japanese Famicom, despite being the exact same piece of programming. It didn't have to run 17% slower, but most companies simply figured that the extra programming effort wasn't worth the investment, because A) that was manpower they could devote to new projects instead, and B) the game would sell roughly the same in Europe whether they fixed the problem or not. And that's if they even understood the problem at all; virtually every development studio was (and generally still is) based out of either Japan or North America, leaving European R&D somewhat less of a priority. (Especially considering the communications technology of the era, compared against the staggeringly powerful worldwide infrastructure we enjoy today.)

So, basically, it sucked. And even when companies did try to adjust for the processor difference, the results were unpredictable. The PAL version of Super Mario Bros., tweaked to hopefully run at NTSC speed, instead accidentally runs slightly faster. PAL Tetris plays the music at the right speed, but moves faster, making for a harder game. PAL Metroid runs slower, but also accidentally plays its title music at a higher pitch than originally intended. The music in PAL Kirby's Adventure plays properly, but the game runs its usual 17% slower, except for Kirby, who runs at proper NTSC speed and is thusly far faster than the game intended. It's fascinating, the lot of it, but understandably frustrating for European audiences.

The rise of digital televisions would eventually render the analog NTSC/PAL divide obsolete, giving the world interlaced and progressive standards (the 'i' and 'p' in modern benchmarks, e.g. "1080p") instead -- but unfortunately for European retro-gaming enthusiasts, the modern re-releases of classic games are still slower. What a world! What a world.

Anyway, in a nutshell: the PAL versions of most NES games run 17% slower than their NTSC counterparts, and the music runs 17% slower as well. That's not to say the pitch is different; both versions of the game are reading the same notes at the same point in the programming, the PAL processor just takes 17% longer to get to each note.

Here are the NTSC and PAL versions of Mega Man II's first Wily Castle theme, as an example:

Weird, right? If you grew up familiar with one version but not the other, the other version just seems really, really strange. ("[T]he first time I heard the Balloon Fight theme in proper speed I was completely freaked out".) So it went, too, with this album I wrote: because I was learning Deflemask on the fly while I composed in it, I didn't realize until very late in production that the program has a toggle switch between NTSC and PAL processing speeds. Ultimately, then, I got to live out a very specific and very funny experience of the NES era: I had been writing exclusively in NTSC the whole time, and I had not once considered how it would behave in PAL, and I was totally blindsided by the effects of the translation.

Some of the PAL versions ended up being pretty darn neat, and the ones that didn't ended up being pretty funny. So please enjoy the complimentary, and complementary, PAL versions of each track, included with this album at no extra charge.

And, uh... and yeah. Please enjoy Manitoba Circuit OST! I think that's everything I wanted to cover about the album, but of course I'm happy to talk about it (as you, uh, could probably tell), so please feel free to email me or comment below if there's anything else you'd like to know.

So That's It?

So that's it, I suppose. Eight years and change after Slurpees and Murder began, this is how Slurpees and Murder ends: a cheerfully esoteric, inexorably weird, hopefully compelling project, the last of what I like to think has been a series of them over the years.

This isn't the last you'll hear of me, certainly; I'll still be chirping about Manitoban matters on Twitter with fair frequency, and I have a suspicion I can be persuaded to contribute for other outlets as well. Speaking of which, I figure I'll start myself another new site eventually; it'll have a different focus to it (obviously, since I don't live in Manitoba any more), and I don't feel any particular hurry on it quite yet, but I'll definitely let you folks know when it's up.

For now, however, this is it. I had left this site on sort of a cliffhanger when I'd moved away for new employment, so I'm glad that I'm finally able to send the site out on a high note -- or rather on a series of them, accompanied by tasty bass triangle-waves and an assortment of eight-bit percussion. There's no finer way to go out, if you ask me!

Finally, then, and formally, then:

This concludes Slurpees and Murder. Thank you for reading.

No comments: