The Sound of the Machines
Mega Drive / Genesis
メガドライブ
Two chips, ten channels, and a design flaw that became a sound.
Yamaha YM2612 + SN76489 10 channels · Sega · 1988
The Chip
The Mega Drive had a split personality — and it made music that sounded like nothing else.
The main sound chip was the Yamaha YM2612, a chip that Yamaha had originally designed for professional synthesizers. It used FM synthesis: no recordings, no compressed audio samples. Pure mathematics, running in real time. Six channels, each producing tones generated by oscillators modifying each other's frequencies. The sound it produced was hard-edged, metallic, intensely physical. When pushed to its limits — maximum feedback on the oscillators — it made sounds that seemed to vibrate with their own energy.
Alongside the YM2612, the Mega Drive also carried a second, much simpler chip: the Texas Instruments SN76489, inherited from Sega's older hardware. Four additional channels: three square waves and one noise generator. Where the YM2612 was complex and mathematically dense, the SN76489 was clean, simple, and slightly clinical. Two completely different sonic personalities, expected to produce one coherent piece of music.
Composers who understood both chips could use the contrast productively. The FM channels handled the body of the music — the brass, the bass, the gritty textures that the YM2612 did naturally. The PSG channels handled the light: high counter-melodies, shimmer, the bright frequencies that FM synthesis tends to round off. The noise channel became hi-hats, grain, texture. Put all of that together and you had ten channels speaking two different languages — made coherent by giving each one a different job.
Channel reference (technical)
Tomita's Track
Song of the Ladder
ラダーの歌
What to listen for
The lead brass in this piece is the YM2612 running at maximum feedback — the setting that produces the chip's characteristic grit and rasp. That abrasive quality is not a mistake. It is the point. This whole piece is built around what happens when you lean into the chip's most aggressive voice rather than trying to smooth it out. Also listen for the two-chip architecture: the FM voices carry the body and the growl, while the SN76489 squares and noise channel provide the high shimmer and the grain of the rhythm. Two different sonic languages agreeing by doing different jobs.
The YM2612's sixth channel can play either FM synthesis or 8-bit PCM drum samples — not both at the same time. So I used FM channels 1 through 5 and all four SN76489 PSG channels, giving the FM voices the body (metal brass, growling bass) and the PSG voices the light and grain (high chords, noise hi-hat). Two different chips, two completely different sonic languages — made to agree by giving each one a different job.
The real story this piece is built on
MD-1: The YM2612 had a design flaw in its output circuit that added subtle distortion to quiet signals — later called the 'ladder effect.' Some composers wrote music that relied on this distortion. When Sega corrected the flaw in the YM3438, music written for the original chip sounded thinner.
How this piece was made
The first time I pushed the YM2612's feedback dial to maximum — value 7, the top — I heard something I did not expect. It was not a musical instrument. It was metal breathing. A creature that does not exist, made of frequencies and anger. Anyone who has played a Mega Drive game knows that sound: the rasp, the grit, the 'jarrr.' That is feedback at maximum. This piece is built around that sound, because this piece is built around the ladder effect — the design flaw in the YM2612's audio circuit that added distortion to quiet sounds. Engineers called it a bug. Composers turned it into a signature. When Sega fixed the bug in a later version of the chip, some music written for the original actually sounded wrong. The flaw had become the flavor. I used maximum feedback for the lead brass so that every note carries that grit — the 'character flaw that became the character.' The SN76489's noise channel handles the hi-hat: pure high-frequency noise, chopped into 8th notes. That combination — FM body, PSG light, noise grain — is the anatomy of Mega Drive sound.
Techniques used in this track
- YM2612 Algorithm 4 feedback brass at maximum feedback (fb=7): intentional distortion — the 'ladder effect' turned into lead voice
- YM2612 Algorithm 0 FM bass: four-operator chain producing a growling low-end line
- YM2612 CH1 and CH2 harmony brass (fb=6): slightly softer, filling the harmonic gap below the lead
- YM2612 FM kick (CH5) + FM snare (CH4): percussion made entirely from frequency modulation
- SN76489 high-chord voice (CH1): the clean PSG tone providing brightness above the FM grit
- SN76489 noise hi-hat (8th-note pulse): the 'grain' of the rhythm — pure noise turned into texture
- 34-bar structure repeated four times, each pass rising in pitch register: building toward a sense of arrival
Three things about the Yamaha YM2612
FM synthesis was already well-known from professional synthesizers when the Mega Drive launched. Yamaha's DX7 keyboard had used the same core technique to reshape pop music in the early 1980s. The difference was context: a DX7 cost several thousand dollars and sat in a recording studio. The Mega Drive cost a fraction of that and sat in a child's bedroom. The same mathematics, a different world — and different compromises, which produced a different, very specific kind of sound.
FM synthesis does not record sounds. It manufactures them from mathematics.
The Yamaha YM2612 — the main sound chip in the Mega Drive — does not play back recorded audio the way the Super Famicom does. It generates tones in real time using a technique called FM synthesis: one sound wave modifies the frequency of another, and the resulting interference patterns produce complex, evolving timbres. The chip had six channels, each built from four oscillators called "operators" that could be connected in eight different configurations called algorithms. Change the algorithm and the same four operators produce completely different sounds — a clean piano note from Algorithm 5, a growling brass stab from Algorithm 4, something that sounds vaguely like a human voice from Algorithm 7. Composers who understood the mathematics well enough could produce almost any sound they could imagine. Those who did not spent a great deal of time producing sounds they had not intended.
A design flaw became a signature. When it was fixed, the music sounded wrong.
The YM2612 had a defect in its audio output circuit. Quiet signals passing through the chip picked up a subtle additional distortion — a kind of rough, slightly gritty texture that audio engineers later named the "ladder effect," after the ladder-shaped circuit topology responsible. In studio equipment, this would have been an unacceptable flaw. In a game console, it became something else. Composers writing for the Mega Drive heard the effect and incorporated it into their work. The grit was part of the voice. When Sega later revised the hardware and produced a corrected version of the chip — the YM3438 — music that had been written specifically for the original chip sounded different when played on the new hardware. Thinner. Less itself. The bug had become the feature, and removing the bug damaged the music.
Channel 6 had to choose: FM instrument or real-sounding drums. Never both.
Six of the Mega Drive's ten audio channels came from the YM2612, and one of those — the sixth — had a special capability. It could switch between two completely different operating modes. In FM mode, it behaved like the other five channels: a full FM instrument. In DAC mode, it could play back raw 8-bit audio samples — drum hits, sound effects, short voice clips. But it could only do one of these things at a time. The choice was permanent for any given moment in the music. Sega's own composers — writing Sonic the Hedgehog, Streets of Rage, other defining Mega Drive titles — typically dedicated channel 6 permanently to DAC mode, playing real-sounding kick and snare samples. The other five FM channels handled everything else. That split — real drum sounds over FM-generated instruments — is a significant part of what makes Mega Drive music feel the way it does: big, physical, surprisingly alive for a machine running on math.
The sound of organized anger
FM synthesis does not approximate acoustic instruments. It creates sounds that belong entirely to the digital world: hard-edged, metallic, with overtones that no acoustic instrument produces naturally. The Mega Drive's music sounds the way it does not because anyone was trying to sound aggressive — but because the chip's natural voice, left to its own character, tends toward intensity.
The composers who understood this best did not fight the chip. They learned its language: operators and algorithms instead of strings and breath. They found the settings that produced the grit, the growl, the gleam, and they used those qualities as the building blocks of something new. Not an imitation of what other music sounded like. Something specific to this machine.
The ladder effect — the "bug" that added distortion to quiet sounds — is a case study in what happens when a flaw becomes a character trait. Engineers called it a defect. Composers heard a voice. When Sega removed the defect, they took the voice with it.
The Mega Drive did not sound gritty because anyone wanted grit. It sounded gritty because the mathematics of FM synthesis, at consumer price points, sounds like that. The composers who heard it clearly and leaned into it made music that still has no equal.
Explore further
Want to hear this chip on original hardware? Browse our tested Mega Drive collection →