Sync beyond a hard reset to A: soft and weak sync, reversing sync, and a second master in one voice #124

New issue

Open

opened 2026-10-05 23:46:12 +00:00 by bcox · 0 comments

bcox commented

2026-10-05 23:46:12 +00:00

Owner

bcox asked on 2026-10-05 for a ticket for each engine gap in his moregaps.md feature table that no existing ticket covers. Engine claims are from reading the tree at f6aa8eba on main. Instrument claims are from that day's research, read mostly as archive.org OCR of the manuals by research agents and not re-checked; the sources are listed in that document. Nothing here is built, measured or listened to.

Summary

Oscillator A is the VA layer's only sync master, and sync is a hard reset of the slave's phase on A's wrap (dsp/va_layer.h:1721-1745). Three things the instruments do are outside that:

Soft or weak sync. The SH-5's sync switch is weak, off, strong; the service notes do not say what weak does electrically. The CEM3340's soft sync forces an early reversal at the triangle's peak rather than a reset. The oscillator spec deferred soft sync.
Reversing sync. The CEM3340's hard sync reverses the triangle's direction rather than resetting it. The accumulator here already runs backward for through-zero FM (dsp/va_oscillator.h:177-193), so a reversal is a direction flip at the master's sub-sample position; but a reversing sync puts corners at arbitrary points in the triangle, and the oscillator leaves triangle corners uncorrected today (:199-200).
A second master. The Mono/Poly's double mode syncs VCO 1 to 2 and VCO 3 to 4 at once. Here that is two layers. The Synthex's manual has one direction (2 to 1); Sound On Sound says each can sync to the other, which the schematic's one SYNC net per voice was not traced far enough to settle.

The work, as a sketch

A sync mode per slave: hard (today), soft (reverse at the master's wrap if the slave is past its own midpoint, else nothing -- the 3340's rule, roughly), reversing (flip direction at the master's wrap, with the PolyBLAMP corner the flip makes).
A sync source per slave rather than a fixed master A, so any lower-indexed oscillator can be a master; lock already names a lower-indexed source per oscillator (dsp/va_layer.h:546-553), so the plumbing has a precedent.
Bandwidth: a chip's sync sidebands never fold; here they are bounded by PolyBLEP and the oversampling the sync already asks for.

Open

What the SH-5's weak sync is. Without a schematic reading it may be soft sync or a partial reset.
Whether reversing sync on a saw or pulse means anything, or is triangle-only as on the chip.

bcox asked on 2026-10-05 for a ticket for each engine gap in his `moregaps.md` feature table that no existing ticket covers. Engine claims are from reading the tree at `f6aa8eba` on `main`. Instrument claims are from that day's research, read mostly as archive.org OCR of the manuals by research agents and not re-checked; the sources are listed in that document. **Nothing here is built, measured or listened to.** ## Summary Oscillator A is the VA layer's only sync master, and sync is a hard reset of the slave's phase on A's wrap (`dsp/va_layer.h:1721-1745`). Three things the instruments do are outside that: - **Soft or weak sync.** The SH-5's sync switch is weak, off, strong; the service notes do not say what weak does electrically. The CEM3340's soft sync forces an early reversal at the triangle's peak rather than a reset. The oscillator spec deferred soft sync. - **Reversing sync.** The CEM3340's hard sync reverses the triangle's direction rather than resetting it. The accumulator here already runs backward for through-zero FM (`dsp/va_oscillator.h:177-193`), so a reversal is a direction flip at the master's sub-sample position; but a reversing sync puts corners at arbitrary points in the triangle, and the oscillator leaves triangle corners uncorrected today (`:199-200`). - **A second master.** The Mono/Poly's double mode syncs VCO 1 to 2 and VCO 3 to 4 at once. Here that is two layers. The Synthex's manual has one direction (2 to 1); Sound On Sound says each can sync to the other, which the schematic's one SYNC net per voice was not traced far enough to settle. ## The work, as a sketch - A sync mode per slave: hard (today), soft (reverse at the master's wrap if the slave is past its own midpoint, else nothing -- the 3340's rule, roughly), reversing (flip direction at the master's wrap, with the PolyBLAMP corner the flip makes). - A sync source per slave rather than a fixed master A, so any lower-indexed oscillator can be a master; lock already names a lower-indexed source per oscillator (`dsp/va_layer.h:546-553`), so the plumbing has a precedent. - Bandwidth: a chip's sync sidebands never fold; here they are bounded by PolyBLEP and the oversampling the sync already asks for. ## Open - What the SH-5's weak sync is. Without a schematic reading it may be soft sync or a partial reset. - Whether reversing sync on a saw or pulse means anything, or is triangle-only as on the chip.