RSS icon Home icon
  • Pilotwings (Nintendo, 12/21/90)

    Posted on March 18th, 2011 keving 4 comments

    Pilotwings Resort is coming out pretty soon, but why not concentrate instead on the only game among the SNES launch titles that got me really, really excited? (What can I say — Mode 7 was really amazing to me, in a way that the PlayStation wasn’t somehow.)

    This game, featuring music by my beloved Soyo Oka (who must have really like that “blaaa” instrument because it’s used in two tracks), is one of several to use a DSP-1 coprocessor in order to speed up the trigonometric calculations required for the quick scaling/rotation seen in-game. F-Zero does not use this coprocessor despite having even faster scaling/rotation moves. This is because — and I forget who told this to me, so I can’t give a source — something like half of the game ROM is composed solely of precalculated cosine tables, obviating the need to come up with the figures in realtime.

    Japanese Wikipedia claims (unsourced) that the first shipment of Pilotwings in Japan did not include a DSP-1, something which I don’t think is actually true. What is true is that the game may have either the DSP-1, DSP-1A, or DSP-1B chip onboard. The 1A is a simple hardware revision to make the chip smaller, while the 1B is the same as the 1 except with a few bugfixes to the microcode that drives the device. You can tell which chip is inside your Pilotwings without opening up the cartridge because the 1B revision actually triggers a bug that’s easily demonstrated. Start up the game and keep it running until you get the gameplay demo with the light plane. If the plane lands correctly, the game’s running on a DSP-1 or 1A; if it crashes well in advance of the runway, you’ve got a DSP-1B.

    Neat, huh? And until I started researching this, I didn’t even realize there was that “secret” side pool you could hit with that one bonus stage.