The clever chaps at XILS Lab have given Roland's classic VP330 a new lease of life as a soft synth.
Comprising a simple vocal synth, an even simpler string synth and a basic vocoder, the Roland VP330 Vocoder Plus has to be one of the most unexpected success stories in the history of electronic musical instruments. I suspect that it left a far greater mark on our industry than even its designers might have hoped, so it's perhaps surprising that we've had to wait until 2013 for a fully fledged emulation to appear from an established manufacturer. But now the XILS Lab V+ has arrived, and this promises to retain all of the character of the original while extending its facilities far beyond what was practical in 1979. Let's find out if it was worth the wait.
When Roland released the VP330, string synthesis was becoming a company speciality, with the RS202 and RS505 ensemble keyboards making inroads into the sales of established instruments such as the Solina. But the Strings section in the VP330 was not what you would call advanced or flexible, with just a single 4' voice, a basic tone control and a monophonic Attack stage that created the 'sucking' sound that you obtain from all paraphonic synths. It should have been ghastly... but it wasn't. If you worked within its limitations, the results were rather pleasing.
To evaluate the Strings section in the V+, I sat my original VP330 alongside the Mac hosting the soft synth, and found that the two could be made to sound very similar. They're not identical, but the differences are small enough for me to be confident that the two would be interchangeable in a mix.
As promised by the company's blurb, the Strings section in the V+ also adds a lot that wasn't found on the VP330. For example, the Waveform slider in the Mixer/Arp panel allows you to change the underlying timbre of its sound from something vaguely sawtooth-ish to something vaguely square-ish, while the Filter slider dramatically alters its timbre. Next to this lies the Ensemble slider, which allows you to control the depth of the ensemble effect, but perhaps of greatest benefit are the Attack and Release tabs that help to overcome the shortcomings of the original's envelopes. The results are still constrained by the weird architecture of applying the envelopes to the outputs from the top-octave divide-down oscillators rather than to each note individually, but that's in keeping with the VP330 itself, so I can't complain.
Looking back, the VP330's Strings section was totally outshone by its Human Voice section, which will forever be associated with the soundtracks from Chariots Of Fire and Blade Runner. Untreated, the underlying waveform and formant filters produced just two 'ahh' sounds, but, with the ensemble effect engaged, the result oozed character. When I saw Tony Banks of Genesis replace his Mellotron with a VP330 on the Duke tour in 1980, I realised that the world as we knew it had ended.
I compared the Male 8', Male 4' and Female 4' voices of the V+ to those of the VP330, with and without ensemble, and with and without vibrato. Again, there were slight differences, particularly in the way that the original synth articulates the start of new notes/chords. But I wouldn't describe the V+ as better or worse than the VP330, just a little different.
As with the Strings, the Mixer/Arp tab reveals a further set of controls for the Human Voices. These allow you to adjust the contribution of the 8' and 4' registers, the frequency and intensity of the vocal formant, and the depth of the ensemble effect. What's more, a Poly Attack tab and the Release tab (which is shared with the Strings) again allow the notes to 'speak' correctly. These controls extend the soft synth's capabilities far beyond the original instrument's, creating all manner of new tones that aficionados are going to love. What's more, the Waveform slider is also shared, allowing you to make (albeit, very subtle) changes to the underlying timbre of the 4' Human Voices although, strangely, not the 8' voices. All in all, this is good stuff.
For many users, the coup de grace of the VP330 was its Vocoder. This offered 10 frequency bands and a high-pass filtered signal path that added back fricatives and sibilants to ensure that the result was intelligible enough for lyrics to be understood following treatment. With an External Synth input as well as a microphone input, it even allowed you to vocode other instruments and things such as music from LPs and cassettes.
I again carried out an A/B comparison. I found that the characters of the two were very similar but, unlike my VP330, the V+'s vocoded sounds contained a significant amount of the unmodulated carrier. I double-checked my conclusions by recording a snippet of speech and used this as the modulator for both instruments. This revealed that the problem was caused, at least in part, by noise; the V+ is much more susceptible to it than the VP330, and almost any amount of broadband noise in the modulator makes the carrier audible (as, indeed, it should). I discussed this with Xavier Oudin at XILS Lab and, within a couple of days, he had supplied a link to the prototype of a revised version. This added a control that allowed me to determine the minimum signal level accepted by each of the analysis bands, and was a huge improvement. At the time of writing, there's still a little tweaking to be done, but the new version will be available to everyone by the time that you read this review.
As you would expect, the V+ vocoder provides everything that you'll find on a VP330, plus a wide range of new facilities that you won't. For example, the Analyse section allows you to determine how quickly the vocoder follows the modulating signal, which is great, and it also offers a tab that creates an underwater effect, which is a gimmick. To the right of this you'll find the Filters section, which allows you to determine the cut-off frequency of the high-pass filter used to eliminate the body of the modulator before adding back any fricatives and sibilants to the outgoing signal. A second fader here also affects the Qs of the filters in the filter bank, which has a marked effect on the nature and intelligibility of the vocoded signal. Finally, there's the Pitch Tracker section. Unlike the VP330, you can make the pitch of the carrier track the pitch of the modulating signal. (Strictly speaking, this shouldn't be in the vocoder section, since the same carrier is used for the Strings and Human Voice, but let's not quibble.) The tracking is far from perfect, and I wouldn't play a melody in this fashion, but the output from the pitch tracker can be used as a modulation source, where its inaccuracies and instabilities can add an interesting degree of variation to the operation of other parameters.
Ah yes, modulation. Like the original, the V+ offers Glide (a slide up to the played pitch, not portamento), pitch bend, and the ability to apply vibrato to the Human Voice and the vocoded sounds. It also offers a powerful modulation matrix with six slots within which you can — subject to sensible limitations — apply 11 sources to 24 destinations with both positive and negative polarity (see table.) Furthermore, the modulation section contains a dedicated, MIDI-sync'able LFO with six mixable waveforms, plus a five-stage (HADSR) contour generator. Nonetheless, I feel that this section is a work in progress because many common routings (say, applying velocity to output level) required the use of two slots to obtain the response I wanted.
There are also three additional effects. XILS Lab describes the phaser as an emulation of an analogue unit but, while its perfectly usable, it lacks the deep 'swooosh' of, say, my Electro-Harmonix units. What's more, it increases the output level by around +7dB, which is not good. Next, there's a simple reverb offering small, medium and large options, with controls for reverb level, reverb time and low-pass damping. Here, the signal level was boosted by as much as +10dB. Finally, there's the Stereo Space effect. Derived from the company's Polykobol soft synth, this allows you to position the outputs from the Strings, the Human Voice and the Vocoder in a two-dimensional space, which can be rather interesting but, again, the output level can be hugely greater than the input level, so handle it with care.
Despite V+'s teething problems, it can sound far superior to sample-based recreations of the VP330, if only because things such as its weird envelopes are to a great degree correctly modelled. It's also hard to overemphasise how flexible the V+ is when compared with the VP330. Take the Strings, for example. You can adjust the waveform and filter to obtain a suitable tone, use the envelope generator to shape the note, add both tremolo and vibrato, control the modulation depth using a mod wheel, aftertouch or a ribbon, and add a touch of Glide to create a solo violin or cello patch that a VP330 couldn't dream of producing. If this sounds a little daunting, the V+ also comes with a wide range of factory sounds residing in XILS Lab's standard patch management system, which allows you to categorise sounds by bank, author, project, and so on, to sort them, compare them to one another, export and import sounds and, well, you know, all the usual stuff.
Then there are things like the arpeggiator. I couldn't think why XILS Lab added this until I heard it used within the Fresh Air factory sound. It's not a facility that springs to mind when I think about string synths, vocal synths, or vocoders but, with its up, down, last note and random modes, a three-octave range, the ability to play 'broken chords' (which I haven't encountered since Grade 5 piano!), a programmable chord playing facility, plus MIDI sync'able rate, swing and Gate time, it's quite a sophisticated package.
Another unexpected bonus is the Hiss control added to the prototype that I received during the review. Earlier versions generated the gentle background noise that's noticeable on the original when playing the Human Voices alone. The new version allows you to choose whether this is present or not, and to what degree. So, if you prefer digital cleanliness (which I do) you can have it. That's a nice touch. In common with the VP330, the V+ also provides a fixed keyboard split that allows you to play different sounds on either side of middle 'C' — say, the violin mentioned above in the upper two octaves, and a vocoded sound with support from the Human Voice in the lower two. Unfortunately, I also have to report a few additional bugs that I discovered during the review. Some — such as the incorrect calibration of the vibrato rate tool-tip — are trivial. Others are not. For example, I was able to create problems while assigning the controllers on my Arturia AE keyboard, ranging from graphical errors to total crashes of the plug-in and the host. Oh yes, and while it's not a bug as such, the manual was not written by a native English speaker, and would benefit from a revision.
The VP330 is a fabulous instrument. It only does three things, but it does them with a style and quality that have never been surpassed. I'm therefore delighted that there's now a soft synth that emulates it. Before testing the V+, I thought that the additional features might detract from the almost perfect simplicity of the original, but I needn't have worried. In fact, I'm sure that I'll end up using it, and that's something that very rarely happens to products that I review. With genuine VP330's now changing hands for £$2000 and more, the V+ is also excellent value. Once the vocoder has been finalised and the remaining bugs have been exterminated, there might be no better way to obtain this important family of classic sounds.
Monophonic Sources Polyphonic Sources Monophonic Destinations Polyphonic Destinations
Modulation wheel Velocity Human formant resonance Level
Pitch-bend MIDI note number Human formant frequency Human Voice attack
Aftertouch Poly pressure Vocoder resonance Strings attack
Vibrato LFO HADSR envelope Vocoder filter bank freq shift Release
Mod section LFO Vocoder release Human lower 8' level
Glide Carrier pitch Human lower 4' level
Pitch tracker Carrier level Human upper 8' level
Tremolo frequency Human upper 4' level
Tremolo level Strings lower level
LFO frequency Strings upper level
LFO level Vocoder lower level
Vocoder upper level
Installing the V+ plug-in on my MacBook Pro was painless once I had replaced the initial release with v1.0.3. There's no stand-alone version, so I installed the VST, AU (Mac only), RTAS and AAX versions. Authorisation was via the eLicenser software and dongle although, for Pro Tools users, iLok is also supported.
Some hosts don't handle plug-in audio effects (which is what the V+ vocoder is) in the same way as plug-in instruments (which is what its Strings and Human Voice are). Consequently, XILS Lab has created two versions; one that supports the Strings and Human Voice alone, and one that adds the vocoder and allows you to push audio through it. I was able to launch both versions in Plogue Bidule and elsewhere without difficulty, although DP7.24 recognised only the instrument version. Reassuringly, XILS Lab have confirmed that both versions work on their DP7 and DP8 systems, so I'm confident that this was just a configuration problem on my Mac.