UVI are offering a very reasonably priced emulation of a classic Moog — but which classic Moog is it?
Since I started playing soft synths about 15 years ago, I’ve noticed two things happening. Firstly, they’ve become better... much better. The imitative ones have gotten closer to emulating the sounds and responses of their inspirations, and the innovative ones have incorporated more ways to generate, twist and play sounds. Secondly, they’ve become boring. Does the world need another Minimoog or Hammond B3 emulation, or yet another way to slice up waveforms, modulate them according to the positions of the planets on your 21st birthday, chop the results into little pieces and then reassemble them according to some algorithm unknown outside the maths departments of major universities? As a consequence, I’ve started looking for products that break new ground, either in terms of recreating something rare and unobtainable, or in terms of introducing new — but understandable and playable — methods of synthesis. The UVI PX Apollo appears to be one of the former, inspired by the legendary Moog Apollo whose only outing was ELP’s Brain Salad Surgery tour of 1974. But is it?
To answer this, we need to look at the history of the Apollo. The first instrument to bear this name was built in 1973 as part of the unreleased Moog Constellation mono/poly/bass synth. The monophonic instrument, the Lyra, was also used by ELP, and bits of it eventually ended up in the Micromoog and Multimoog. The bass synth also survived in part as the Taurus pedals. But the polyphonic bit in the middle, the four-octave, single-oscillator-per-note instrument described by Moog as “a true polyphonic synth and electric piano” was never seen or heard again, although much of its technology later appeared in the Polymoog Keyboard, which was released in 1975.
The Polymoog was much enhanced when compared with the Apollo; the single oscillator per note had become two oscillators per note, a sustain phase had been added to its contour generators, and its keyboard had been extended so that it could be played more like a piano. Unfortunately, this made it expensive and unreliable so, in 1978, the powers-that-be at Norlin (who owned the Moog brand) decided to release a cheaper, simpler, preset version that lacked its synthesis controls. They then transferred the name ‘Polymoog Keyboard’ to the new instrument, and rebranded the original the ‘Polymoog Synthesizer’. To this day, this creates confusion that could have been avoided had they chosen a new name for the new instrument, and it seems that they might have done so but it was dropped before the product was released. If an instrument now owned by the Moog Foundation is genuine, there was a prototype of the 1978 Polymoog Keyboard that, apart from the name plate, looks all but identical to the released product — and it’s called the Moog Apollo.
What makes this all the more confusing is that both versions of the Apollo still exist. Although no longer working, Keith Emerson’s instrument can now be found at the Audities museum in Canada, while the later one is pretty much functional and in 2014 was even seen on stage in Moog’s home city of Asheville, North Carolina. So, which of these instruments forms the basis of the UVI PX Apollo? Despite UVI’s references to “the oft veiled world of unique, fringe and unreleased electronic and acoustic instruments” and to “a larger project that was eventually abandoned”, there’s no confusion. Its preset sounds and its ability to sustain notes reveal that this is a soft synth based upon the Polymoog Keyboard, not the original Apollo.
You can access UVI’s samples of the Moog’s raw sounds by selecting the 00-Init Mono preset and then stepping through the buttons in the Master Voice Selector. These are not always accurate in the way that they respond when played, but the underlying timbres are authentic (even down to the spurious, although not intrusive, background noises) and, in some ways, playing them with no modification other than perhaps a touch of chorus and reverb is this soft synth’s strongest suit. Indeed, had I designed it, I might have been tempted to leave things at that and restrict the editing system to those controls found on the original. However, the majority of potential users will be relieved to find that I didn’t, and that it offers far more by way of sound generation and sculpting.
In total, it comprises three sound generators called Bass, Osc A and Osc B, which can be selected individually or in any combination. The first is an emulation of the Polymoog Keyboard’s simple bass sound, complete with its dedicated low-pass filter and volume control. When selected, this is shown on the representation of the keyboard in the GUI as extending from MIDI note E2 downward although, in fact, it extends from Eb2 downward. This has no effect upon the useability of the bass section, but it strikes me as something that should have been noticed and fixed.
Next, Osc A is the sample replay engine based upon the Polymoog’s presets. These have been recorded monophonically, which means that you can’t adjust the detuning as you could on the original synth; you have to accept the settings used when UVI recorded them. To its right, Osc B is a completely different beastie, offering 12 digitally generated waveforms derived from UVI’s Falcon synthesis workstation. These include analogue-style waves such as triangle, saw, ramp, and three flavours of pulse, together with a small selection of more complex waves. Almost all of the PX Apollo’s editing capabilities can be applied independently to Osc A and Osc B by depressing the large A or B buttons that sit alongside them, which means that (overlooking the Bass section) it includes two independent polysynths that are mixed at the effects/output stage.
After the master volume and pan faders, the first group of editing controls comprise a four-stage contour generator applied to the audio amplifier. It isn’t faithful to the Polymoog Synthesizer, which generated an ADSD amplitude contour rather than the ADSR presented here, but it’s close enough to allow you to shape sounds in the same way. The amount of contour applied is velocity sensitive, and there’s also a button marked Vel>A that shortens the attack time when you hit the keys faster, which is a real bonus when emulating acoustic instruments. Strangely, the Vel button in this section works backwards. When up (off) the amplifier is velocity sensitive; when depressed (on) it’s not. This means that it’s working as a defeat button, which is fine, but not in keeping with the rest of the synth.
Much greater differences are revealed when you turn to the resonant filter section. Unlike any version of the Polymoog, this offers three modes (low-pass, band-pass and high-pass) and a second polyphonic ADSR contour generator. Ignoring the Polymod cards that generated its initial voices, there was only a single filter and its associated contour generator in the Polymoog, so any sounds shaped by this were paraphonic, and held notes were retriggered when you played new ones. Since the PX Apollo filter section is truly polyphonic, each note is now shaped correctly; this isn’t authentic but it’s a big improvement. However, the Vel knob (which one would imagine would connect velocity to the filter cutoff frequency or the contour amount) appears to affect the input or output level, ie. the loudness. If there’s any effect upon the cutoff frequency or contour amount it’s minimal, and I can’t imagine that this is what UVI intended.
The second edit page allows you to determine the pitches of the oscillator sections, and offers a polyphonic envelope that either decays from a high pitch down to the normal pitch of each note, or rises from a low pitch up to the normal pitch of each note, with slew times ranging from almost instantaneous to approximately two seconds. (In the distant past, this was sometimes called Slalom.) You’ll also find a button marked Mono here, which converts PX Apollo into a monosynth, and the Slalom effect then becomes a conventional portamento.
The next panel allows you to ‘double’ the Polymoog sounds, either sending alternate notes to the left and right channels, or spreading and detuning them to create richer timbres. When Osc A is selected, there’s an additional control here called Color, which UVI explain “shifts colour based on adjacent samples”. Initially I had no idea what this meant, but experimentation revealed that — depending upon the position of the fader — it causes each note to be played using pitch-shifted samples from one or more semitones above (or below) the one that you’ve played, thus creating a deeper tone or a degree of munchkinisation. It’s an interesting enhancement, but not as useful as you might wish because many of the sustained samples have inherent modulation that becomes unnaturally fast if you raise the fader too far, while the percussive sounds can decay too quickly. Nonetheless, it’s a neat idea. If Osc B is selected, you also have the option to select the number of voices — from one to eight — in the unison as well as the amount of detune and spread, allowing you to create effects ranging from a warm chorus to the unavoidable swarm of angry wasps at a college garden party.
The third panel on this page determines whether the modulation wheel affects the independent LFOs generating pitch modulation (vibrato) and loudness modulation (tremolo), or the filter cutoff frequency, or any combination of these. Strangely, the vibrato rate is shown as a number, while the tremolo rate is shown as a MIDI Clock division, even though the movement of the slider isn’t quantised that coarsely. I can’t see why this should be so but, whatever the reason, both LFOs were running freely when tested.
The third page contains a more sophisticated LFO offering sine, triangle and square waveshapes as well as sample and hold (S&H). This runs independently of the LFOs on the previous page, and you can select whether it’s retriggered each time that a new note is played, and whether the LFOs affecting each note are synchronised with one another or not. You can direct the results to the pitches of the notes, their loudnesses, and their individual filter cutoff frequencies. Above this, the upper panel allows you to craft your own delayed LFO waveforms using 16 steps that you can smooth (or not) as you please. You can then direct this to the loudness and cutoff frequency of each note.
There’s a slider in each of these LFO panels marked Osc B PWM. This applies something akin to pulse-width modulation to the Osc B waveforms but, since all the waveforms are affected (not just pulse waves), and the documentation describes them as Phase Distortion waves, it might have been more accurate to label the sliders Osc B PD. Unfortunately, the centre point of the modulation doesn’t seem to be in quite the right place so, rather than a smooth up/down/up/down sweep in the timbre, you obtain a up / down / pause / up / down / pause... and so on. Something else to be tweaked in an update, I think.
Moving on again, we come to the FX page. This contains an overdrive and four additional effects: chorus, phaser, delay, and reverb. I would advise caution when using the overdrive since there is no low-pass filter or speaker emulator following it to soften the results, so it can sometimes be a bit bright and harsh. The other effects do what they promise, and although I would be tempted to use external chorus and phasing effects — whether within the UVI Workstation host (see 'Hosting PX Apollo' box later) or from elsewhere — for top-quality results, I was surprised by how useful I found the delay and the reverb.
Finally, on the ARP page, there is a pair of arpeggiators, one each for the Osc A and Osc B sections. These offer up, down and up/down modes covering up to three octaves, with user-definable gate length, and are married to a pair of 16-step sequencers that are hard-wired to the velocities of the notes being played. The results can be surprisingly expressive, but I managed to freeze Osc B ‘on’ with its arpeggiator running on a number of occasions, and only removing the soft synth from its slot within the UVI Workstation would silence it.
I had installed PX Apollo in the hope that, for the first time, I would be able to investigate something closely related to the original Moog Apollo from 1973. Finding that the sampled part of its engine is based upon the Polymoog Keyboard from 1978 and that its editing system is both extensive and unrelated to either the Apollo, the Polymoog Keyboard or the Polymoog Synthesizer was a disappointment. Nonetheless, the fact that I could play the raw sounds from the Polymoog Keyboard (albeit without the ability to adjust the detune of its two oscillator banks) was pleasing, especially since I sold mine many years ago and no longer have the stamina to own and maintain one.
The software comes with a good selection of factory sounds, most of which layer the Osc A and Osc B sections in interesting ways. They’ve been divided into 12 categories — Voice (based upon the Polymoog presets), Arpeggio, Bell, Brass, Bass, Chords, Keyboard, Lead, Pad, Pluck, Strings, and Sweeps — and it’s well worth stepping though these to explore what’s possible. Since Osc A and Osc B have such different characters, I wasn’t certain how well they would work with one another to create coherent sounds, but I needn’t have worried. In fact, trying things such as sprinkling arpeggiated PPG-ish fairy dust from the digital waveforms over a deep pad from the Polymoog sounds is really simple and can sound gorgeous. I found that I had to increase the buffer size from 128 to 512 samples to avoid glitching with complex patches, but that was a small price to pay. However, the useful ranges of some of the controls — particularly the modulation amounts — tend to be in too narrow a band around zero, which means that it can be tricky to dial in just a hint of something using your mouse or touchpad.
Since I couldn’t find a fine-movement mode (and the manual makes no mention of one) I taught the controls to respond to MIDI CCs from the Arturia master keyboard that I was using to play the synth. This was also simple: I just right-clicked on the desired control and then twiddled the desired knob, fader or switch on the Arturia, and then found that it was much easier to dial in small values, which improved the experience considerably. It also meant that I could automate PX Apollo from any host system capable of recording and transmitting MIDI CCs while playing. I was even able to ‘Learn’ the 14 Polymoog presets and the 12 Falcon waveforms, assigning these to the Snapshot buttons on the master keyboard, which was a pleasant surprise. Unfortunately, later selection of these almost always caused PX Apollo and the UVI Workstation host to crash, which was not. What’s more, I could find no way to save the assignments in PX Apollo itself; it was necessary to save a Multi in the UVI Workstation to do so.
Before finishing, I have to comment on the errors and omissions in the manual. For example, it states that you can direct the Mod page LFO to the drive parameter, but you can’t. And an explanation such as “Filter: Vel — Set the filter’s velocity sensitivity” leads me to ask, “set the velocity to affect what aspect of the filter? The input level, the resonance, the cutoff frequency, the amount of contour applied to one or all of these...?” Another sentence refers to “filter depth” when it’s referring to the filter cutoff frequency. The manual also calls the Bass section a sub-oscillator, which of course it’s not. A rewrite and proof read wouldn’t go amiss.
I asked for the opportunity to review PX Apollo because I was fascinated by the possibility of playing something that emulated the original synth from 1973 and, even after this hope was dashed, I was still keen to see whether it would replace my long-departed 1978 Polymoog Keyboard. In truth, it did this surprisingly well, and the additional bonuses of Osc B, the extensive modulation capabilities and the effects made it a far more flexible synth that I had expected. But it’s vital that UVI implement a better testing regime to iron out the bugs and oversights; these aren’t arcane or difficult to find, so I’m at a loss to understand how they have been allowed to reach the final product. The good news is that, once they’ve been eradicated, PX Apollo has the potential to be a nice product that offers an interesting alternative to other soft synths.
You can’t run PX Apollo as a straightforward AU, VST or AAX plug-in; you have to run it within UVI’s Workstation (currently v2.6.5) or Falcon synthesis hybrid engine (currently v1.0.5). This shouldn’t be a problem because the company provide the Workstation free of charge as a host, and it is this that appears in your plug-in lists. It can host numerous UVI synths simultaneously in a huge Multi and apply an extensive library of effects to them so, once you start to use it, you may find that this is itself a valuable addition to your software toolbox.
In the main text, I questioned whether the second version of the Moog Apollo is truly a Polymoog Keyboard prototype. One of my reasons for doing so is a consequence of the 14 preset sounds that it hosts. In part 1 of his seminal A-Z Of Analogue Synthesizers, Peter Forrest noted that some of the earliest Polymoog Keyboards had the presets shown in the first column below. It seems that this configuration was short lived, and the first three presets were soon replaced with the Vox Humana, String 1 and String 2 presets found in all subsequent instruments. So here’s the weird thing... The voices on the Moog Apollo are the same as the second revision. If it were a prototype, why would Moog’s designers change its 14 signature sounds when launching the product, only to change them back again shortly afterward? Maybe we’ll never know but, whatever the truth is, it’s good that the PX Apollo is based upon the latter set, because the Vox Humana was for many people the best reason to own and maintain the instrument.
|Preset number||Polymoog Keyboard v1 presets||Polymoog Keyboard v2 presets||Apollo & PX Apollo presets|
|1||String Solo||Vox Humana||Vox Humana|
|2||String Chorus||String 1||String 1|
|3||String Funk||String 2||String 2|