XILS-lab MemoryTone Review: What Matters for Bass

Three Memorymoog Emulations in One Year

If you've been anywhere near synth-plugin news in 2026, you've noticed a pattern: everybody suddenly wants to sell you a Memorymoog. Arturia shipped Memory V in May. Cherry Audio followed with Memorymode 2 at the end of August. And on September 9, XILS-lab threw its hat in the ring with MemoryTone, priced at $79 as an intro offer (rising to $199 on November 1, 2026). Synth Anatomy's own headline for the launch put it bluntly: "the year of the Memorymoog emulations isn't yet over."

That's a lot of competing software chasing one 1982 polysynth that was notorious for glitching on stage. So the real question for anyone reading this instead of a spec sheet is simple: does any of this matter if you're an Ableton producer who mostly cares about low end? The short answer is yes — the original Memorymoog earned its reputation partly because of its unison mode, which could stack all of a voice's oscillators onto a single note for what's often described as "a fat 18-oscillator monophonic bass and lead sound." That's the feature worth paying attention to across all three plugins, not the polyphony specs that dominate the press releases.

What MemoryTone Actually Is

XILS-lab's MemoryTone models the Memorymoog's three-oscillator-per-voice architecture across 16 voices (18 oscillators total), with a self-oscillating 24dB/octave ladder filter, two ADSR envelopes, six modulation slots, and six built-in effects (chorus, delay, phaser, reverb, EQ, and a resonator) with flexible routing. It ships with MPE support, a mono/poly glide mode, and — this is the part that matters for bass work — "a very massive unison" mode for stacking voices into one thick low note.

One detail that's easy to miss in the spec lists: the third oscillator per voice can drop roughly five octaves below normal range and detach from keyboard tracking entirely, which is normally used as a modulation source but doubles as a way to reinforce a fundamental well below where the other two oscillators are sitting. XILS-lab has also built the emulation around modeling the original hardware's oscillator interactions, voice-to-voice variation, and filter behavior specifically — not just copying the basic signal path — which is the kind of detail that tends to separate a synth that sounds "close enough" from one that actually behaves unpredictably in the way analog gear does.

Formats are VST, VST3, AU, and AAX on both Mac and Windows.

How the Three Stack Up

None of these plugins are identical takes on the same hardware, and the differences actually matter if you're deciding which one to spend money on for bass work rather than pad and lead sounds.

Arturia's Memory V is the most conservative of the three — it leans hard on faithfully modeling the original circuit rather than adding new architecture, and its 6-to-12 voice range with a drive control makes it a solid choice if you want the closest thing to the real hardware's character without a lot of modern bolt-ons.

Cherry Audio went the opposite direction with Memorymode 2: it's not just an emulation, it's a redesign. The dual-layer architecture lets you run two independent 16-voice Memorymoogs simultaneously, split or layered, each with its own filter and sequencer — and the up-to-48-voices-per-layer unison stacking is a genuinely different tool than the original hardware ever offered. If you want to stack an absurd amount of oscillator mass under a single bass note, this is the one built for that specifically, and its lower up-front price ($69, or $39 if you already own the first Memorymode) makes it the easiest one to justify on a whim.

MemoryTone sits in between: XILS-lab's reputation is built on obsessive circuit modeling (their past releases have leaned heavily on getting the nonlinear, slightly-unstable character of analog gear right rather than smoothing it out), and MemoryTone's approach — massive unison plus a de-tunable sub oscillator — reads like a synth aimed specifically at people who want a huge, slightly-unruly low end rather than a pristine one.

What This Means If You're Shaping Bass in Ableton

None of these plugins replace what you do downstream, but the choice does change your starting point. A synth like Memorymode 2, with its 48-voice-per-layer unison stacking, is going to hand you an enormous, already-wide low end — which means you'll likely want to tame the stereo image before it hits your mix bus, since a bass that's that wide rarely translates well in mono playback systems. MemoryTone's approach, with the detachable sub oscillator running well below the other two, is closer to a synth-generated version of a classic sub-layering trick: you're getting fundamental reinforcement built into the patch instead of stacking a separate sine underneath in your DAW.

Whichever one you land on, the same principles that apply to Ableton's own bass-focused devices apply here too. If you've read our breakdown of Corpus for bass texture or Shifter for bass sound design, the logic carries over directly: a huge unison patch from any of these three synths is a starting point, not a finished bass sound. High-pass everything except the fundamental before it hits a return, decide early whether the patch needs to collapse to mono below a certain frequency, and don't be afraid to route the raw oscillator mass through something that adds harmonic content rather than just more low-frequency energy — a wall of identically-tuned oscillators can sound huge in solo and disappear in a full mix if there's nothing distinguishing it harmonically from the rest of your low end.

Is Any of This Worth Buying Right Now?

If you don't own a Memorymoog emulation yet and bass patches are your priority, Cherry Audio's Memorymode 2 is the easiest recommendation on pure value — $69, dual-layer architecture, and unison stacking built for exactly this use case. Arturia's Memory V is the pick if authenticity to the original hardware's behavior matters more to you than modern extras, though at €149 it's the most expensive of the three for what is, by design, the most conservative feature set.

MemoryTone's intro price of $79 makes it a reasonable third option, especially if XILS-lab's reputation for accurate modeling matters to you more than raw voice count — but it's worth remembering that price rises to $199 on November 1, 2026, so the window to grab it at the lower price is limited. None of these are essential purchases if your current bass chain already gets you where you need to go; they're worth it specifically if you've been missing a genuinely fat, stackable analog-style low end in your plugin folder and want one built for it rather than repurposed from a pad synth.

Takeaway: three real companies released three real Memorymoog emulations in 2026, and the unison-stacking feature that made the original hardware a bass monster is the actual differentiator between them — not the voice counts or the price tags that lead most of the coverage.