VGM Entry 07: Other chip options
(Thanks to Tish at FFShrine for the banner)
I am at the unfortunate disadvantage of having no clue what key terms such as “FM Synthesis” and “Programmable Sound Generator” really mean, and no amount of reading technical explanations or listening to arbitrary examples of audio employing one or the other is really going to fill me in. I feel like it is the very sort of thing this series of articles is intended to explain, but it’s not currently within my grasp.
One thing I’d like to know is what makes arcade games like Tube Panic (Nichibutsu/Fujitek, 1984) sound so much better than their arcade predecessors of only a year or two prior. (Unfortunately this composer’s name has eluded me, perhaps lost in translation.) This game uses a General Instrument AY-3-8910 chip, or so I am told, which is a PSG. So did Jungle Hunt, and the two are worlds apart. Jungle Hunt‘s three very basic tones could barely hold themselves together, constantly breaking out of rhythm and sounding quite primitive even when they all synced up. Of course the glitchiness was part of the charm, but Tube Panic is an entirely different animal. There is definitely no sense that the system is struggling to contain the music, and the tones are much fuller. What changed? And if it’s the case that later arcade games stacked multiple audio chips where early ones did not, how exactly does this effect the end product?
There is one thing I’ve noticed, and it’s probably both an amateur observation for those who know what they’re talking about and a pointless one for those who don’t. But it seems to me like audio employing FM-synthesis is much cleaner.
Thexder (Game Arts, 1985) for instance was composed by Hibiki Godai and released for the NEC PC-8801 the same year that this system began to incorporate a Yamaha YM2203 sound chip, which, as best I understand it, used FM synthesis. Whatever that actually entails, what I seem to be hearing here is a lack of distortion never attained with the AY-3-8910, or with the Commodore 64 SID for that matter (another PSG). That’s not necessarily a good or bad thing–distortion was the perfecting touch to the early Ultima soundtracks (the Mockingboard also employed multiple AY-3-8910 chips) and it would be the focal point for some of the best ZX Spectrum titles. But there is a noticeable difference in clarity, and if I had to guess I’d say it’s the dominant difference between FM synthesis chips and PSGs.
The most impressive early consequence of this cleaner sound is Marble Madness (Atari, 1984), which used the YM2151, an FM synthesis chip similar to the YM2203. The music Brad Fuller and Hal Cannon manage to create here is gorgeous and completely unbecoming of an otherwise conceptually mundane video game. The music of Marble Madness can essentially function as a stand alone semi-ambient synth album. With a few exceptions and a little longer content it could have been commercially released independent of any game to reasonable acclaim, and it is not all that particularly different from the sort of works you might expect on the Yamaha keyboards employed by 1980s synth musicians. Tasking Brad Fuller and Hal Cannon with the job and providing them with the sound chip to get it done might have been one of the only things Atari did right in the 1980s.
The last thing to note here is that Earl Vickers is credited as the Marble Madness sound programmer. This is one of the earliest games for which I’ve noticed different names associated with ‘composition’ and ‘sound programming’, and it’s a confusing distinction which will impact plenty of future discussions.