Arturia Memory V

Arturia’s software version of the Memorymoog may be better than the original...

Ah, the Memorymoog. Simultaneously a wonderful synthesizer and a pile of poo. The former because, on a good day, it can sound glorious. The latter because, if you haven’t switched one on for a while, you’ll probably have to open it up and spend an afternoon poking around inside to perform an arcane hardware tuning routine, just to be able to play the bloody thing. Sure, all Memorymoogs are now more than 40 years old, so it’s a miracle that they work at all, but I never have these problems with a Jupiter‑8, an OB‑8 or even a Prophet 5 Rev 3 (let’s not talk about my Rev 1) so I can love and hate the Memorymoog simultaneously. There has to be a simpler way to obtain the sounds of the original without all of the farting around... which brings us to Arturia’s latest offering.

The Technology

Memory V is a 12‑note polyphonic, velocity‑ and aftertouch‑sensitive emulation and enhancement of the Memorymoog that you can run as a standalone application or as a VST, AU or AAX plug‑in. As you would expect, it follows Arturia’s tried and tested pattern, with the original synth’s voicing married to an Advanced section that offers extra modulators, macros and a powerful four‑part polyrhythmic arpeggiator, together with settings for MIDI, sync and MPE.

Starting with the on‑screen synthesizer itself, the oscillators are laid out in the same fashion as before, albeit with the octave buttons replaced by rotary selectors. The only extra facility here is a button to hard‑sync osc 1 to osc 3, which may not be something that you’ll use a lot, but is welcome nonetheless. The mixer layout is identical, and seeks to emulate what is, perhaps, the most notable characteristic of the Memorymoog; at any value above 5 or thereabouts, the input into the filter is overdriven with a mild, soft‑clipping distortion. If you set up Memory V appropriately, you can see this in the oscilloscope display toward the left of the panel.

The signal now passes through the filter and amplifier sections. Here, the functional differences from the original are more marked. For example, a button alongside the cutoff frequency knob switches between the model of the Memorymoog’s 24dB/oct response and a new 12dB/oct mode, while a second button alongside the Emphasis (more commonly called Resonance or ‘Q’) knob engages what Arturia call ‘bass compensation’, emulating a different type of architecture in which the filter doesn’t attenuate the signal below the cutoff frequency when you increase the Emphasis. In addition, the Memorymoog’s four keyboard tracking amounts have been replaced by a knob offering a smooth range from zero to 100 percent. All of these are valuable additions.

At a glance (and notwithstanding different legends for the contour times and name changes for the buttons that determine how they work) the ADSR contour generators for the filter and amplifier appear to be the same as before. But there are two significant differences. Firstly, the A, D and R ranges have been extended. Secondly, the button that enabled the release stage on the original has been replaced by one called Velocity Env Times, which links the contour times to velocity — hit a note harder, and its A, D and R stages are longer. This is in addition to sending velocity to the amplifier gain and filter cutoff frequency in the pane on the left of the synth.

At the end of the visible signal path, the three volume controls of the original (master, patch and headphones) have been replaced by a master volume, pan/unison spread, and drive, the last of which would more logically appear between the oscillator and filter sections because it determines the overall level into the latter of these. But this isn’t the end of the real signal path because, tucked away in the Advanced panel, there’s a four‑slot effects section that draws upon 18 major types of effect, some of which can sound excellent.

A factory patch making full use of the arpeggiator (shown here), effects and other Advanced facilities.

Modulation is provided by an enhanced, global LFO Modulation panel (now with sync and retriggering) and a per‑voice Voice Modulation panel, but again with the use of space‑saving rotary selectors. Both sections offer two additional destinations: volume and pan.

Three slots in the Advanced panel provide access to the additional contouring, a function generator, a random generator, a voice modulator and a modulation sequencer, all of which are ridiculously flexible and can control great swathes of the voice generation and effects parameters. Then there are Arturia’s usual macros, which allow you to group multiple parameter changes under a single controller. If you want to use (for example) a knob on your MIDI controller to alter the pitch of one of the oscillators, the filter cutoff frequency and the feedback of a flanger simultaneously, each with a different response, this is where you do it.

Behind a discrete panel, you’ll find seven knobs that allow you to determine how much various aspects of the sound differ from one voice to the next. (Arturia call this ‘dispersion’.) If you turn all of these to zero, you obtain a slightly sterile ideal of the Memorymoog sound. But if you turn all of them — and the Vintage knob in the Global panel — to their maxima, you’ll get some idea of the vagaries of the original. I suspect that a Memorymoog Plus or a Linntronics Memorymoog would behave in a more gentlemanly fashion than my early model, but I still found that dispersion could capture much of its character without making it so undisciplined as to be unplayable.

Despite its myriad additional capabilities, I would venture that Memory V is simpler and more intuitive to use than the Memorymoog itself. In large part, this is because the original’s System Controller was neither simple nor intuitive. Replacing the Moog’s 100 memories (identified merely as 00 to 99) with Arturia’s advanced patch storage and selection system is a no‑brainer, and Arturia’s undo/redo when programming is always welcome.

In Use

It was now time to see whether the sound of Memory V could live up to Arturia’s claims. I started by comparing all of the initial waveforms against those generated by my Memorymoog. In every case, they were all but indistinguishable, although there’s an ongoing phase issue when you mix multiple oscillators in perfect tune. I then created some filter sweeps using various waveforms and different amounts of resonance. Again, the results were excellent, although I could tell which synth was which by the variations in the voices. So I adjusted the dispersion and increased the value of the Vintage knob even further. At around 75 percent of maximum, I found myself within a gnat’s wotsit of the original.

I then created some useful sounds, concentrating on the Memorymoog’s strengths of synth strings, polybrass, pads, basses and unison leads. In all cases, I was able to recreate its sound on Memory V or achieve something so close as to make no difference. Inevitably, there are cases where the hard synth and the soft synth diverge from one another but, as I’ve said before, if you spend your time looking for these, you’re not making music.

Imagine what an enhanced Memorymoog might have been like had the original Moog Music not gone out of business. It might have sounded much like this.

Having satisfied myself that I could emulate the Memorymoog, I proceeded to create patches that took advantage of velocity and pressure sensitivity. You may feel that this isn’t in the spirit of the original, but I loved the results. Then I unleashed the full capabilities of the soft synth, adding all manner of additional modulation paths and effects. The results were impressive. Imagine what an enhanced Memorymoog might have been like had the original Moog Music not gone out of business. It might have sounded much like this.

Conclusions

Memory V echoes many other modern soft synths. When programmed sympathetically it can sound like its inspiration, but it offers far more than the original, and I can only scratch the surface of the enhancements in a review of this length. Of course, it doesn’t feel like standing behind a Memorymoog, and nobody is going to come up to you after a gig and beg to see your laptop. But if it’s the sound that matters, that shouldn’t be an issue. Inevitably, some will claim that it’s nothing like the real thing but, if you calibrate and tune a Memorymoog as best you can and then set up the dispersion within Memory V to emulate the remaining inconsistencies, you’ll find that their characters can be remarkably similar.

When I liberated my Memorymoog from storage to perform these tests, it tuned two voices and sounded awful. After a few days it tuned five, but Voice 1 was so far out that the software tuning algorithm was never able to make it useable. I couldn’t be bothered to spend an afternoon calibrating the damn thing, so I just defeated the recalcitrant voice and conducted this review using a 5‑voice Memorymoog. In contrast, Memory V fired up instantly, offered me 12‑note polyphony and could sound pretty much the same. Oh yes... and it costs around one percent of the price of a second‑hand original. Which would you choose?