The original Moog Model D had no dedicated LFO at all, and the reissue only added one in 2016, forty-six years after launch. Players got vibrato by switching oscillator three to its Lo range, releasing it from the keyboard so it ran free, then feeding it into the modulation mix. A workaround, treated as a feature ever since.
The rest of Oracle's source instruments are barely more generous. The Korg MS-20 has a single modulation generator with two outputs. The Sequential Pro-One has one main modulator, wired to two destinations. The SH-101 runs one LFO between 0.1Hz and 30Hz. The Sonic Six is the outlier with two, and the DFAM has no LFO section on its panel whatsoever.
So the modulation system in the plugin is the one part of it that has no vintage counterpart. Six LFOs per patch, any of them assignable to anything, with a custom shape editor on top. That is not a recreation of anything. It is a different instrument bolted onto old waveforms.
The constraint did the sound design for you. With one source, you had to decide what the patch was about, and the answer was almost always a slow filter movement or a touch of vibrato, because those are the two things that make a monosynth sound alive. Nobody was modulating six parameters at once, so nobody made patches that fell apart in a mix.
Now you get dragging a modulation source onto a knob and repeating it five more times in under a minute. That is the trade you accept with Oracle's stack of six assignable LFOs, and it explains something about big preset libraries generally. Presets have to sell themselves in an eight-second audition. Movement sells. So the modulation tends to sit heavier than anything you would have patched yourself.
That mismatch is a large part of why old recordings feel handmade. The wobble slides against the drums, arrives slightly early, then slightly late. Two takes of the same patch are never identical.
Some of these machines could be clocked, but not in the way people assume. The SH-101 has an external clock input, and on the Pro-One the sequencer and arpeggiator take their timing from the LFO frequency, so the modulation drove the timing rather than the other way around. The tail wagging the dog, and it sounded like it.
A host-synced LFO inverts that completely. Lock a rate to a sixteenth note and every sweep lands on the grid, every take is identical, and the part sits perfectly still inside the arrangement. Cleaner. Also flatter, in a way that is hard to name until you hear the two side by side.
Start by asking what the patch is for. A bass part needs one thing moving, usually filter cutoff, sometimes nothing. A pad can carry two or three, provided they run at different rates and land on different destinations, because two LFOs at the same speed hitting related parameters just make one bigger LFO with a phasing problem.
Then unsync something. Take the slowest source in the patch, pull it off the tempo grid, and set it to a rate that has no clean relationship to the tempo, roughly 0.2Hz to 0.5Hz for a pad. The part stops repeating on a bar boundary and starts breathing, which is most of what people mean when they say a sound feels analog.
Custom shapes are worth more than extra sources here. A slightly asymmetric ramp, or a shape with a flat section in the middle, does more for a patch than two more sine LFOs, because the ear notices contour before it notices count. The MS-20's modulation generator worked on exactly this principle, offering shape control on a single source rather than a second source with fixed waveforms.
And when a preset arrives sounding busy, open its modulation page before you touch anything else. Disable sources one at a time and listen to what each removal costs. In a fair number of cases you will find two or three doing nothing you would miss, which tells you the patch was built to demo rather than built to sit under a vocal.
The rest of Oracle's source instruments are barely more generous. The Korg MS-20 has a single modulation generator with two outputs. The Sequential Pro-One has one main modulator, wired to two destinations. The SH-101 runs one LFO between 0.1Hz and 30Hz. The Sonic Six is the outlier with two, and the DFAM has no LFO section on its panel whatsoever.
So the modulation system in the plugin is the one part of it that has no vintage counterpart. Six LFOs per patch, any of them assignable to anything, with a custom shape editor on top. That is not a recreation of anything. It is a different instrument bolted onto old waveforms.
One modulator and two destinations bought a lot of discipline
Look at what the Pro-One actually let you do. One modulation source, routed through a small matrix, landing on oscillator pitch or filter cutoff. That was the menu.The constraint did the sound design for you. With one source, you had to decide what the patch was about, and the answer was almost always a slow filter movement or a touch of vibrato, because those are the two things that make a monosynth sound alive. Nobody was modulating six parameters at once, so nobody made patches that fell apart in a mix.
Now you get dragging a modulation source onto a knob and repeating it five more times in under a minute. That is the trade you accept with Oracle's stack of six assignable LFOs, and it explains something about big preset libraries generally. Presets have to sell themselves in an eight-second audition. Movement sells. So the modulation tends to sit heavier than anything you would have patched yourself.
Tempo sync removes the drift you were paying for
An analog LFO free-runs. It is an oscillator with a slow rate control and no notion of where bar three starts, which means a two-second filter sweep never lines up with the same beat twice across a take.That mismatch is a large part of why old recordings feel handmade. The wobble slides against the drums, arrives slightly early, then slightly late. Two takes of the same patch are never identical.
Some of these machines could be clocked, but not in the way people assume. The SH-101 has an external clock input, and on the Pro-One the sequencer and arpeggiator take their timing from the LFO frequency, so the modulation drove the timing rather than the other way around. The tail wagging the dog, and it sounded like it.
A host-synced LFO inverts that completely. Lock a rate to a sixteenth note and every sweep lands on the grid, every take is identical, and the part sits perfectly still inside the arrangement. Cleaner. Also flatter, in a way that is hard to name until you hear the two side by side.
Restraint is the setting the hardware used to enforce
The fix is not to avoid the modulation section. It is to spend your six sources the way someone with one source would have spent theirs.Start by asking what the patch is for. A bass part needs one thing moving, usually filter cutoff, sometimes nothing. A pad can carry two or three, provided they run at different rates and land on different destinations, because two LFOs at the same speed hitting related parameters just make one bigger LFO with a phasing problem.
Then unsync something. Take the slowest source in the patch, pull it off the tempo grid, and set it to a rate that has no clean relationship to the tempo, roughly 0.2Hz to 0.5Hz for a pad. The part stops repeating on a bar boundary and starts breathing, which is most of what people mean when they say a sound feels analog.
Custom shapes are worth more than extra sources here. A slightly asymmetric ramp, or a shape with a flat section in the middle, does more for a patch than two more sine LFOs, because the ear notices contour before it notices count. The MS-20's modulation generator worked on exactly this principle, offering shape control on a single source rather than a second source with fixed waveforms.
And when a preset arrives sounding busy, open its modulation page before you touch anything else. Disable sources one at a time and listen to what each removal costs. In a fair number of cases you will find two or three doing nothing you would miss, which tells you the patch was built to demo rather than built to sit under a vocal.
