Synthesizer
Reviews : Keyboard
I've been in this game a long time, but I can't remember the last time that a synthesizer raised passions as high as does the Q. You only have to read the various synth-related Internet-based forums and the letters written by Q supporters to realise that Waldorfophiles have adopted something akin to a siege mentality: it's us against the capitalist pigs at Korg, Roland, and Yamaha, chaps! Aux armes! To your word processors and web browsers! We will not be unheard!
If this seems over-dramatic, you haven't seen the opprobrium I received following my preview of the Q (see SOS May '99). I felt that I had given the instrument as it then stood a generous thumbs-up, concluding with the statement that "if the finished Q lives up to this sort of promise, it's going to be one hell of synthesizer." As a reward, the following question appeared upon the SOS Readers' forum, "What was Gordon Reid smoking when he previewed the Q?" This was followed by, "It's too bad Gordon Reid can't recognise a classic when it's right in front of his face." Not that I give a damn, but it demonstrates the passion with which Waldorf's footsoldiers will defend their chosen synthesizer. And indeed, anything that commands this degree of support deserves closer scrutiny, so now that the Q has come on somewhat, let's get stuck in.
When Is A Released Synth A Finished Synth?
If you're wondering what I just meant by 'the instrument as it then stood' and 'now that the Q has come on somewhat', I'd better explain. The Q used for this review is the same physical synth that I previewed, so all the differences lie in the version of the operating system loaded. This makes a big difference, however. The May SOS preview was originally planned as a full review, but -- even ignoring the bugs and crashes -- the operating system at that point was too unfinished to make this realistic. This was despite my upgrading of the instrument to OS v1.07 before I completed the preview. It wasn't until September that Waldorf announced OS v1.12, a version that makes reality of many of the
|
By the time I had downloaded and loaded Q OS v1.12, Waldorf had already posted v1.13 to their FTP site, so I had to repeat the exercise of upgrading a couple of days later (and it's not necessarily as easy a task as you might think -- see the 'Upgrade Downer' box for more on this). Anyway... both Waldorf and I are finally of the opinion that the latest OS gives you a Q that's worthy of detailed scrutiny, so it's time to put the thus-far troubled story of this big yellow beastie behind us and, finally, review the instrument itself.
User Interface & Oscillators
As you'll know if you read the preview in May (and if you haven't, you should, as this review is written to be read in conjunction with it), the Q boasts no fewer than 58 knobs scattered across its control panel (59 if you include the large red data-entry knob), 40 buttons, and a fair sprinkling of LEDs. It is this as much as anything else that endears it so to the analogue (and modelled analogue) synth anoraks out there. Nevertheless, there would be many more controls were it not for the Shift button. This provides a second, related function for each control. But even with the Shift system, there are still a considerable number of functions that you can only access using the tiny screen and the editing system. There's another way in which the Q's user interface, despite its high number of physical controls, fails to please, but I'll return to this point later.
The Q is a 'physically modelled analogue' synth similar in concept to the Novation Supernova, Clavia Nord Lead, Access Virus, and Roland JP8000. Each of these uses powerful DSPs to emulate th
e action and idiosyncrasies of analogue circuits. The Q's three audio oscillators each offer the full complement of standard analogue waveforms -- pulse, sawtooth, triangle and sine -- but it's not this that makes them remarkable, it's the unsurpassed range of octaves available: everything from an unheard of (and unheard) 128' to a migraine-inducing 1/2'. These are complemented by semitone and detune controls that -- to a resolution of approximately 1 percent of a semitone -- allow you to program any pitch across the entire frequency range.
The waveforms are genuine recreations of analogue-type waves. In other words, the purity of digital sines, squares and saws has been eschewed in favour of something that sounds more like the waveforms produced by a genuine analogue synth. However, unlike many synths with analogue oscillators, the Q won't let you select more than one of its digitally simulated waveforms at any given time. Instead, it offers independent PWM and independent FM of all three oscillators, plus oscillator sync of Osc 2 by Osc 3. Add to this the dozen or so pitch-modulation sources buried in the menus, variable key-tracking, pitch-bend and portamento, and you have a powerful package of features. Unfortunately, I found that the lower settings were rather too coarse for gentle modulation. I frequently wanted settings between, say, 1 and 2, whereas the range of values from about 20 up to 127 was unnecessary except for creating silly noises. Waldorf really ought to recalibrate this in a more logarithmic fashion so that the greatest amount of control is concentrated where it is most meaningful. Oh yes, and what of the two oscillator positions marked Alt 1 and Alt 2 which I mentioned in my preview? They are still not implemented, nor have any public decisions on their eventual uses been made at the time of writing. Three steps forward, two steps back!
The Mixer
The output from all three oscillators, together with a Noise source (which also doubles as an external input) and a Ring Modulator (RM), passes to the Mixer. In Waldorf's limited preliminary Q manual the RM appears to have three inputs, but when tested, I found that it has two: Osc 1 is the Carrier and Osc 2 the Modulator, so all is well. Indeed, I think that this is an excellent emulation of a ring modulator, and I consider it to be one of the Q's strongest features.
Waldorf have implemented the external input in an interesting way. A stereo jack socket on the rear panel, allows you to insert two signals that the Q treats as independent sound sources. The menu system then allows you to select which two of four options (Noise, Ext L, Ext R and the summed Ext L+R) you balance in the Mixer. Unfortunately, and like so much else in the Q, it is buggy. If you use an external sound as a source in the Mixer, you can in principle patch the filters, amplifiers and envelopes to allow that sound through (treated, of course, by the filters and amplifiers) when you press the keys. But when you do so, this happens: first note, fine; second note; the result sounds out of phase; third note, fine; fourth note, silence! If you play subsequent notes, the cycle simply repeats. Just as SOS was going to press, news appeared from Waldorf indicating that there was a good chance that v1.18 of the operating system might fix this, but they also stated that in the event the software upgrade failed to cure the problem, a factory hardware tweak would be necessary... groan!
The Filters
The Q offers a novel form of signal routing to its filters. Each input has an associated Pan control in the Mixer (which Waldorf call Balance) that determines how much of the signal passes directly to Filter 1 or Filter 2 (the routing diagram can be seen on the Q's front panel -- see the pic on page 98). The Routing knob then allows you to configure the filters in series or in parallel, or any percentage between. This is very powerful, particularly since you can modulate the routing.
There are 10 resonant filter types altogether, plus bypass:
24dB/octave low-pass.
12dB/octave low-pass.
24dB/octave high-pass.
12dB/octave high-pass.
24dB/octave band-pass.
12dB/octave band-pass.
24dB/octave notch.
12dB/octave notch.
Comb+.
Comb-.
The combination of these and the Routing makes many extreme filtering options possible. For example, two 24dB/octave LP filters in series provide a 48dB/octave filter, and two 12dB/octave band-pass filters in parallel offer formant filtering with its many possibilities for creating vocal sounds.
Beyond mere filter type and Routing, the Q also offers Drive Level, key-tracking, cutoff frequency modulation, and filter FM. I'm at a loss to explain the difference between the last two of these, and they work in similar fashion (although the results are not identical). All of these facilities are duplicated for Filter 1 and Filter 2 so you can, in effect, create bi-timbral patches without using the Multis (of which more in a moment).
Oh yes... I imagine that all the analogue anoraks out there will want to know more about the character of the filters. Fair enough. They are clean and precise, so they sound more 'digital' than 'analogue' (whatever you take that to mean). You can sweep and tweak them like analogue filters, and they are almost entirely free of the stepping that plagued early DSP synths -- which is nice. And for the benefit of the Waldorf fan club, who have made some extravagant claims for the similarity of the sound of the Q's filters to various synths of yore -- I have placed the Q alongside Prophets, Oberheims and various other classic analogue polysynths, and I have no doubt that the Q is different. Not better, not worse, but definitely different.
As for the filter resonance, at high settings (120+) the filters self-oscillate nicely, and it was while I was experimenting with this that I came up with some of my favourite Q sounds: clean, pure, and very musical (with key-tracking of 100 percent you can play these as you would the oscillators). As you reduce the resonance, the self-oscillation diminishes and becomes unstable until, at a setting of precisely 113, it disappears. Indeed, the transition between 113 and 114 is so precise that the Q is not always capable of those sounds on the absolute edge of self-oscillation that are so fashionable in the late '90s. Given the difficulties involved in modelling analogue filters, this is not altogether surprising. Not to worry... I found that all manner of excellent resonant sounds were available, and I particularly liked those that used tuned noise as an element.
Before moving on, I should give special mention to the Q's Comb Filters. Or rather, I would like to, but every time I used them, the Q crashed and I had to disconnect it from the mains to bring it back to life. This is a shame, because I love comb filters and, for the few moments I had access to them, the Q's sounded interesting. Unfortunately, every time I tried to program a detailed sound using them, silence told me that all was defunct in Deutsche-DSP-land.
Amplifiers & Effects
The Amplifier offers just two knobs and these, in typical Q fashion, control four parameters. These are the amplifier modulation amount and source, and the overall volume and velocity sensitivity. So let's move on to... the Effects sections.
From the front panel, everything looks positive enough. There are two effects units in series (FX1 and FX2), but these offer just five effects -- Chorus, Flanger, Phaser, Delay and Distort -- and fewer parameters than 1970s stomp boxes did. This would be OK if the effects sounded excellent, but they don't. No matter how you program them, they lack depth. That's not to say that the Q can't produce lush sounds in the same way as chorus-less Prophets or Oberheims. It can. But the beautiful chorused strings of analogue ensembles and Roland polysynths are quite beyond it. Similarly, the warm phasing, flanging and echo effects of my old '70s units are a world apart. If you think that this is a harsh judgement, consider this: I worked my way through all the factory sounds, and the vast majority left both effects units in Bypass. Make of this what you will.
Envelopes, LFOs, & Modulation Matrix
The Q offers four envelopes, called Filter, Amp, 3 and 4. In normal use, these are standard ADSRs, with the first two assigned permanently to the Filters and Amplifier. However, a flick of the Mode control converts them into 6-stage contours that Waldorf describe as ADS1DS2R. As well as offering six stages, this mode also allows you to control the Attack Level, which makes the envelopes much more powerful than they would otherwise be. Unfortunately, the envelope looping promised in Waldorf's documentation has not appeared. Maybe it will in a future update?
The three LFOs are also well specified. They offer six wave shapes (including sample and hold), plus sync and delay, and they will stray up into the audio range if requested. A single edit menu for each LFO also offers key-tracking, fade-in, and various clocking options.
You can assign all the envelopes and LFOs, plus a wide selection of other modulators, as sources within the Q's modulation matrix (see 'The Matrix' box for a full list). There are 16 slots within the matrix, and these are divided into eight Fast Slots (for rapid modulations that must be calculated at audio frequencies) plus eight Standard Slots (for modulations that do not require as much DSP power). This is ample, and one of the high points of the Q's design.
Xphorm
If the patch structure was all there was to the Q, I am sure that I would be a big fan. In fact, there's much more to the instrument than this, with advanced features such as a Multi mode, an arpeggiator, a sequencer, and 'Xphorm'. You'd expect these to really get the pulse going, and they do, but sadly not always in the way that Waldorf would hope!
Xphorm is a morphing feature that allows you to transform the parameters of one sound linearly into the parameters of another, using either key pressure or the mod wheel to control the amount of transformation applied. It works, but the results are not always reliable. Furthermore (although this is not a fault) the pitches of the oscillators are transformed too, so if these are not the same in the source and destination patches, the sound is pitch-shifted as well as tonally modified. Sometimes this can lead to happy accidents, but most of the time it does not, so its use requires care.
I discovered a number of bugs when I was experimenting with Xphorm -- the mode sometimes jumped to Off, and there were occasional hanging notes. Waldorf do admit that 'this feature is under construction' -- but maybe they should simply have withheld the software until it was reliable.
|
Multi Mode
This is, of course, the Q's multitimbral mode, and one of the most significant advances in the Q's functionality since the time of my preview (when it only worked monotimbrally). Each Multi hosts up to four patches (and the plan is to allow you to access up to four Multis at once, though I never got this to work fully). In Multi 1, these are referred to as t
|
I will now try to explain the Multis' effects structure. I say 'try' because the Q has one unlike any other synth I've owned, played, or reviewed! Let's start with the simple stuff. Each of the first patches in the Multis (1.1, 2.1, 3.1 and so on) has just three output options: Main, Sub1 and Sub2. As you would expect, each of these settings allows you to route the first patches through their individual FX1 and FX2 structures to the selected output pair (see the 'Way Out' box for more on these). No surprises here, then.
You might then expect the second, third and fourth patches in each Multi to offer the same arrangement -- but this is not the case. Although each of these subsidiary patches (12 in all) appears to retain its effects from the monotimbral patch structure, it does not. Instead, each offers four additional output assignments: FX1, FX2, FX3 and FX4. These allow you to divert the un-effected output from each patch -- no matter which Multi holds it -- to the Effects structure assigned to one of the '.1' patches -- 1.1, 2.1, 3.1 or 4.1.
This means, for example, that you could have patch 1.1 with distortion and chorus, with patches 1.2, 2.3, 3.2 and 4.3 also passing through these effects. Simultaneously, patch 2.1 could have flange and chorus (with, of course, different parameters) additionally accessed by patches 1.3, 1.4, 3.3 and 4.4, and so on.
Unfortunately, you can't control the amount of effect applied to each of the subsidiary patches -- it's all or nothing -- so this is only a poor man's multitimbral effects structure. But it works reasonably well... once you've worked out what the heck is happening!
Indeed, this is as good a time as any to tell you that much about the Q's operation is arcane. For example, to save a Multi you must first ensure that you are in Multi mode with the Multi setting On, then hit Shift/Store, then name the Multi, and then hit Shift/Store a second time to save. This, of course, is not documented anywhere. Aargh! Surely Waldorf could have designed it to be simpler than this? Indeed, if I hadn't spent days experimenting with every button-push imaginable in Multi mode (and living with the inevitable crashes that occurred every few minutes) you wouldn't be reading this.
Oh yes, and before we move on, let's not forget that the Multi mode has appeared at the expense of the drum synthesizer promised in the Q's early documentation. In other words... there are still no drums on board.
The Arpeggiator
I'm happy to say that the Q's arpeggiator is simple, straightforward, and doesn't crash. There are three modes: On, One Shot, and Hold -- and the top-panel controls allow you to set the arpeggio's range (in octaves), the duration of the note Ons and the tempo. Delving into the menus offers a wider range of options. There are 16 patterns (one of which is user-programmable) and four choices of direction, velocity modes, and parameters relating to ordering of notes, swing, accent, and glide for each step.
If I have to find faults in the arpeggiator, they are just small ones. Firstly, there's no Random setting. This was always one of my favourite features on the Jupiter 8, and I don't know why other manufacturers overlook it. Secondly, the arpeggiator will generate MIDI Clock (although I guarantee you'll never find the command without help) but does not spit out MIDI note information -- shame.
The Sequencer & Global Functions
If the arpeggiator is praiseworthy, I find the step sequencer quite the opposite. Not because the idea of an analogue-style sequencer i
"I have placed the Q
alongside Prophets, Oberheims and various other classic analogue
polysynths, and I have no doubt that the Q is different."
|
Basic step entry is simple enough although, without a manual, it takes a long time to work it out. First you have to enable the sequencer using the so-called Grab Control (obvious, huh?) and then play the desired notes or chords while holding down each of the Step keys. You have to be careful which Bank and Step you're addressing, but it works. You can also add pauses, and hold notes over more than a single step.
Once you have entered your notes, you can set the length of each step and the durations of the notes within the step. You can even rotate the steps to which the notes are assigned, which might be useful if you have a number of patterns linked together. Since a sequence is local to a multitimbral part, you can (in theory) have up to 16 sequences running simultaneously. More usefully, you can have fewer sequences and play other patches over the top. Commendably, the sequencer generates MIDI so you can use it to 'play' external sound sources, and it will work with the arpeggiator.
So far, so good. Now for the bad points. Firstly, the user interface is a mess. Even in OS v1.13, which
|
Finally, we come to the Global menu. This, as always, is a bit of a housekeeping mode, containing parameters for MIDI channel, SysEx ID, local on/off, velocity and pressure curves, footswitch controls, and so on. The only parameter worthy of real mention is the MIDI Clock. This has a neat 'Auto' mode that makes the Q lock to external MIDI Clock when it receives it, or generate clock when it does not. I like this.
In Use
Physically, there's nothing too remarkable about the Q (apart from its yellow colour, of course -- and it's now also available in a less eye-frazzling midnight blue for those who'd rather it blended sympathetically with their studio colour schemes). It's based around a five-octave keyboard that is velocity-, pressure-, and release-velocity-sensitive. I particularly like the way that the chassis protrudes a good half-inch beyond the ends of the keys. In fact, I like the chassis full stop. It's heavy, and it's put together like a Panzer tank. If I dropped the Q on my wooden floor, I know which would come off worse (and it's not the Q).
However, there's one aspect of using the Q which annoyed me no end, and it took me three months to work out what it was: for all this synth's much-vaunted knobs and buttons, it gives you no visual feedback about the parameter values making up the sound. This is because its knobs are infinite rotary encoders (ie. there are no end stops) -- and they are completely uncalibrated. As a result, you look at the panel and, whether the sound is an electric piano, a screaming lead patch, or the softest pad in the world, you get the same information: nothing! For example, if you find a particular sound bright, it could be because the filter cutoff frequency is too high, or the envelope is having a significant effect... or there could be all sorts of reasons buried in the modulation matrix. But there is no way of seeing from a glance at the Q's controls how the parameters
|
Nevertheless, and despite my reservations, by the end of the review period I was creating some killer patches (see the 'Patching It Up' box on page 99). I was also starting to appreciate the Q's clarity and noise-free nature, and making the most of Multi mode as it currently stands. Unfortunately, this was when the number of crashes started to become untenable. Since the only way to cure a crashed Q is to disconnect the power lead from the back of the instrument, this became a source of major irritation. Multis, Xphorms, Sequences... about the only two things that don't crash the Q are basic patches and the arpeggiator. The extent of the problem becomes even clearer when you read posts from Q owners on the Net and the Waldorf User Forum proudly announcing that their machine hasn't crashed in a couple of hours!
Conclusions
It's obvious what's wrong with the Q. Incomplete menus, inoperative controls, constant crashes... it still isn't properly finished. But I have had it with me for nearly six months, and have been able (despite lacking a definitive manual) to assess its progress from v1.02 to v1.13. As a result, I think I can now see where the Q is heading, and why some people will be prepared to overlook the problems. The ethical question remains whether purchasers should be guinea pigs to this extent; I'll let you decide about the rights and wrongs of that.
In addition, not everybody will have the same time and willingness to experiment that I did. If I were a professional musician scraping a living from my music this would be impossible. Indeed, most professionals have no choice but to place a high premium on the speed and simplicity with which they can get acceptable results from their equipment. Presented to someone with a deadline to meet, the Q (which promises much, but only delivers with reluctance) would be a disaster.
Many readers will think that I've been hard on the Q, and fans will complain that I have omitted to mention their favourite feature or use. This may be true, but I have also overlooked a number of bugs that I discovered, so let's call it quits. To deflect some of the
|
As a result, and try as I might, I can't fall in love with the Q. Sure, it's a powerful synthesizer, capable of a wide range of excellent sounds, and sometimes I program a patch or Multi that's simply delightful. Then I try to develop it, and the thing crashes. Consequently, I'm left with the feeling that the Q will remain the preserve of a handful of fanatics. When you consider that Korg and Yamaha ship their instruments by the thousand, and Waldorf do so by the score, this is hardly surprising. But it's not just a question of numbers. The Q does one thing -- analogue modelling. For the same money, you could buy no fewer than four Yamaha AN1xs! Furthermore, three Access Viruses would, in my opinion, give you smoother filters and a more realistic 'analogue' sound. Or you could buy two Korg Z1s, each with their 13 physical models, more powerful effects, and more powerful... well, almost everything else. And then there are the Novation Supernova and Nova to consider, the Clavia Nords, and a bunch of others. At best, the Q has got a hard life ahead of it.
Published in SOS December 1999
No comments:
Post a Comment