SSG is the AY-3-8910 family of square wave / noise generators.
MewFM’s SSG uses a fractional accumulator for square-wave generation, so it moves smoothly even at high pitches, and it has its own AHDSR envelope, SoftLFO, and per-voice filter.
Tone / Noise / Hardware Envelope
| Section | Description |
|---|---|
| Tone | Square-wave generation. The period is set by the MIDI note. Toggle output with Tone Enable. |
| Noise | 5-bit linear noise. Shared by all SSG channels on the chip (matching the original hardware). Toggle output with Noise Enable. |
| HW Envelope | Hardware envelope. Shared by all SSG channels on the chip. Pick Shape (0–15) and Period (1–65535). |
| Level | A fixed volume from 0–15, or, per channel, route the HW envelope to volume. |
About bipolar synthesis
The real SSG outputs unipolar (0..level), but MewFM synthesizes it as a bipolar ±level square wave. To compensate for the resulting doubled RMS, MewFM applies ×0.5 to the final output.
Why SSG sounds louder than FM
Even though FM’s peak amplitude can be higher, perceptually SSG cuts through more.
This is because square waves carry odd-numbered harmonics across the whole spectrum, so at the same RMS they feel +6 to +9 dB louder. PC-98-era players ran into the same thing and manually attenuated SSG to about 0.5–0.7×.
To balance things, use the per-channel Volume slider or the DAW’s Channel Volume CC (7).