The Novation Circuit has been one of my favorite new pieces of gear in a long time, which is impressive considering all the wonderful new gear out there these days. It’s almost perfectly designed for what it is: both a curl-up-on-the-couch music sketchpad and a serious sequencer for playing electronic music. This isn’t a review, but I’ll just say that I’ve found it to be very intuitive to use, capable in both its sounds and sequences, and just a pleasure to hold and use. You should all get one. My only real problem with the Circuit is that I want more: I’d like to make it the centerpiece of my non-computer setup and I need it to do a little more, since I don’t feel like buying four of them. So, this is an open letter to the geniuses at Novation, with a few requests for changes to the Circuit, and then a big request that the Circuit get a big brother. Let’s call it the Circuit Pro.

Here’s my Circuit wishlist (in priority order):
* Ability to edit a single pattern while multiple patterns are playing (i.e. editing doesn’t follow the playing pattern)
* Make note gate time shorter than a single step
* Allow a macro knob to be assigned to oscillator waveform in the patch editor
* Squeeze in a couple more drum tracks
– Assign a midi controller number to each macro in the patch editor, send knob movements as controller data
– SH-101/JX-3p sequencer-style programming (i.e. you play notes and they get entered in one step at a time, with rests and ties possible). Ideally, be able to do this while the Circuit is playing
– Open firmware a la the Launchpad Pro
– Switch modes between 2 synths/4 drums to 0/8 or 4/0 configurations
– Another synth LFO

Here’s what I want in the Circuit Pro:
* 16 tracks in total, 12 drum and 4 synth
* 16 knobs, so that level, sends, etc can be 1-to-1 with the tracks
* 16 macro controls available in synth programs
* Everything from my Circuit wishlist of course
* More sample memory
* Load samples directly with USB drag-n-drop; SD card slot
– More patterns and/or sessions
– More control of drum sound parameters (ideally editable like the synths)
– 2-4 additional audio outputs
– More wavetables in synth oscillators
– Play loaded samples in oscillators
– Finer control of effects settings, maybe a second delay effect
– A couple of analog trigger outs linked to drum tracks

WOULD TOTALLY BUY.

Introduction

I remember when the Morpheus came out, I read the descriptions and thought it sounded fascinating, though a bit of fiddling with one in a store was completely underwhelming. Since then I’ve occasionally wondered:

• How does the Morpheus really sound?
• Can it do things no other synth can do?
• Do later “Z-plane” synths carry on its essence?
• Is it worth having one today?

Luckily, I got to borrow one for a few weeks, enough to get some decent answers, though I don’t claim to be a Morpheus expert. Here are my basic observations, which I’ll go into in more detail below.

• It may have the worst presets ever created. They are almost uniformly muddy, cheezy, boring, and/or unrepresentative. They provide essentially no information about the sound of the synth.
• The annoying button-mashing edit UI isn’t really that bad. Sure, it’s not great, but it’s at least linear and the synth is not that complex. If you otherwise like the Morpheus, the UI isn’t really going to hold you back.
• For such a sophisticated, nerdy synth, it’s surprisingly limited. The set of things you can modulate, particularly in real time, is limited. The handles you have for tweaking the wondrous 14-pole, Z-cube filters are disappointingly few.
• There are truly some interesting options among the 197 filters, though more than half I don’t see ever using, a number of the rest I’d use only rarely.
• I honestly don’t think anyone, even E-mu, has really created filters like this before or since. You could try to recreate them in a virtual modular, but it will
  get very complicated.

If you want the spoiler: in my opinion, while the Morpheus has some interesting, unique capabilities, enough to inspire some serious sound design, it’s too frustrating overall due to its limitations. If E-mu had updated the Morpheus (Morpheus-2000!) with more power and flexibility, it could have been an amazing synth, a true classic. I doubt I’ll ever get one of my own (unless I find a real bargain), but I’m already taking inspiration from its filters into virtual modular designs. Now on to the details.

Demo

First, let’s get right to a demo. This is a very basic patch that runs a sawtooth through a sampling of the filter types available. I’m tweaking knobs mapped to the three main filter parameters: morph, frequency tracking, and transform2. On a few of the patches, I’m also modulating the morph position with a short decay envelope and/or a random note-on S&H. Each filter goes for 16 measures (30-35 seconds).

Architecture

The Morpheus has a fairly typical rompler architecture: each patch is composed of two parts (“primary” and “secondary”; most romplers seem to go with four), each consisting of a sample oscillator, a filter, and an amp, modulated by envelopes and LFOs. In most romplers, these parts are almost completely separate, but in the Morpheus they’re a little more integrated, sharing LFOs and aux envelope (independent amp envelopes) and a single modulation matrix. The Morpheus also offers two “function generators” per patch, essentially an 8-stage envelope, with some crazy extras. One caveat: the LFOs seem to always retrigger at note-on, ruling out slow sweeping filters on fast arpeggiated synth lines. Furthermore, they don’t seem to do the “random value at note on” trick properly either. Aside from the standards, there are four MIDI CC sources available, set globally. Many modulation destinations can be for the primary part, the secondary part, or both. There are plenty of modulation slots, but the matrix is somewhat limited in that some destinations (particularly filter params) can’t be modulated in real time, only at note on. I generally like E-mu’s realtime/note-on distinction, because it makes it very easy to do some nice per-note effects, but of course I want more real-time options. The Morpheus is mostly comparable to romplers of the time, though a little less complex simply in number of layers and programmable parameters.

Oscillators

The samples are pretty good. For non-romplery synthesis, there’s a decent collection of basic waveforms and synth waves. Given the existence of the UltraProteus, I think they could have just tossed most of the real instrument samples and put more synth waves in, but it’s fine. The Morpheus does offer a few wave options you don’t often see: you can turn looping on or off, reverse the wave, and offset the sample start (though you can’t modulate it). You can also “double + detune” each part, which I assume plays a second copy of the same wave, detuned (halving the polyphony), which is handy.

Filters

But of course we’re here for the filters. The basic idea is that, with its 14-pole filter dealie, the Morpheus can offer all kind of filters from simple 1/2/4-pole low/high/band-pass filters through flangers, comb filters, and EQs to complex networks of bandpass filters tuned to formant frequencies. Furthermore, the Morpheus can morph seamlessly between these different filter types. The most Morpheusy of the filters involves a “cube” where three dimensions of modulation can morph among the 8 filters at the “corners” of the “cube”. In theory, this is pretty great, but in practice the Morpheus doesn’t give you control over all of this. Each of its 197 filter settings is an assignment of various filter types to the corners of that cube, and all you control is those three morph dimensions (called, confusingly, “morph”, “frequency tracking”, and “transform 2”). That is, if you’re morphing between a low-pass filter and a high-pass filter, you don’t get independent control over the frequency, resonance, or slope of each of those filters, just the morph position between them. Typically, in a setup like this, the second dimension might control cutoff frequency while the third controls resonance; but not necessarily. Many of the 197 don’t use all of these dimensions — some of the 8 corners are identical – and in many more the corners are fairly similar so that one or two of those dimensions don’t change that much.

The final blow is that only one of these dimensions (“morph”) can be modulated in real time; the other two are note-on only. This means that you can’t, say, hold down a note and turn knobs to morph that sustained note all over the place, which pretty much kills the “I’m going to make an ambient track with one finger like on the Wavestation” vibe. In fact, you get the most satisfaction out of sending the thing a rapid arpeggio run, where you can tweak all three filter dimensions at note-on with knobs to get some nicely varying sounds (though you can’t do slow LFO sweeps). My assumption is that by the time they implemented these insane filters they had no processing power left over to compute realtime modulation, which is why a modern revival of the Morpheus could be so great. Too bad E-mu is no longer around.

Sound

Again, don’t even pay attention to the presets. Many of them are kind of weak and muddy-sounding, which does the Morpheus a disservice. Depending on the wave and filter settings, the Morpheus can sound quite strong and full, though of course many of its more distinctive sounds are weird vowel-like or strangely filtered sounds that you wouldn’t use for techno-bass or anything. It should fit nicely in a mix. However beware: some of the filters vary wildly across their parameter range and you can blow a speaker. Many of the especially peaky filters have a sweet spot where just the right combination of settings causes an explosion. It’s fun in practice but can be a pain if you’re trying to do some musical modulation.

Conclusion

To me, the main question about the Morpheus today is: does it do something I want that nothing else can do? Surprisingly, no one in this era of cheap digital processing power seems to have really copied the Morpheus’s raw filter capabilities. The closest hardware is E’mu’s Proteus-2000 synth and Ultra sampler families, which both include a limited set of Z-plane filters and limited or no morphing (the manuals are unclear, and I haven’t had a chance to demo either recently). E-mu’s X3 or other software seems to have advanced filters, but it’s all seemingly obsolete and hard to buy (or authorize). The only other option is really to assemble your own morphing filters using modular or virtual modular synths. While it’s not too hard to build a nice morphing filter in something like MAX, copying the full morphing topology is a pretty big project; I intend to give it a try but don’t know if I’ll succeed (or, if I do, whether it will sound good or be very usable). Granted, just because nothing reproduces the full Z-plane morphing filters of the Morpheus doesn’t mean that you can’t get pretty close; I suspect that the Proteus-2000 (or other modern synths with lots and lots of filter options) can get most of the way there for most people. But still, the Morpheus (and its sibling the Ultraproteus) is apparently still unique.

For me, even having the Morpheus for a short time, I got frustrated with my inability to modulate it the way I wanted. If I owned one, I can imagine it occasionally showing up for some interesting filter effect, but not being something I would turn to most of the time, which doesn’t seem to make it worth the prices they usually command. The concept though is inspiring — aside from wanting to revisit the Proteus-2000 or Ultra, I’ve been working on some Morpheus-inspired filter ideas in MAX and hope to eventually create my own Morpheus homage.

Favorite Filters

With 197 different filter types and several dimensions of variation, just dialing up a good sound can be daunting on the Morpheus. One of the first things I did when I got my hands on one was set up knobs mapped to morph, frequency tracking, and transform2 and work my way through all the filters, noting down the ones that I liked and thought I might use. Here they are, organized loosely by type. This is, obviously, totally subjective.

Standard
F044 BrickWalLP.4
F045 BrickWal LP2
F046 MdQ 2PoleLP
F047 HiQ 2PoleLP
F048 MdQ 4PoleLP
F049 HiQ 4PoleLP
F050 2poleLoQLP4
F051 4 PoleLoQ.4
F052 4PoleMidQ.4
F054 LowPassPlus
F060 HighAccent.4
F061 HiPassSweep.4
F069 VarSlope.4
F077 PZ Notch
F078 Band-aid
F079 LowQHiQ
F098 EZ Rhodez4
F136 V>FcQuad.4
F142 Comb/HP.4
F144 Cavatate.4
F145 GentleRZ4
F161 HiEndQ.4
F167 SynthWow4
F170 MdlySweep4
F186 MellowPeaks
F189 MildQPole

Flanging
F004 CubeFlanger
F006 Flange 4.4
F007 Flange 5
F011 BriteFlnge.4
F012 Flng>Flng1
F013 O>Flng2
F016 Flng>Flng5
F157 Intervallc4

Vocal
F022 AEParaVowel
F025 AUParaVow.4
F029 Vocal Cube
F030 C1-6 Harms4
F035 Ee-Yi.4
F036 Ii-Yi.4
F041 YahYahs.4
F042 YoYo.4
F043 VowelSpace

Peaky
F065 Rev Peaks
F066 Notcher 2.4
F068 Odd>+
F081 WaWa
F090 LoVelTrum
F092 ShakuFilter
F100 MoogVocSwp
F122 1.5/3KNBPR4
F143 Swirly
F148 SKWEEZIT
F149 Lo/High4
F168 CntrySweep4
F171 StrongShimr
F190 Bonk>CO
F193 Separator

EQ
F073 BassDrumEQ
F074 Snare LPEQ2
F075 HiHatLPEQ

Complex
F084 0>Muter
F088 BrassSwell
F155 HighsTwist4
F175 Harmo
F181 0>Odds
F182 Comb Voices
F184 OddHrm+rez
F185 EvnHrm+rez
F187 AHmBnd.4

Strings
F101 StrngThing4
F102 StrSweep.4

Distortion
F196 ApDistB6.4