U-He Hive Review

When U-He dropped a public beta version of its streamlined, workflow-efficient subtractive synth Hive back in 2014, the excitement for a completely new plugin from U-He was at a fever pitch, even though my enthusiasm for Hive was subdued by my love affair with Zebra 2 and, of course, anticipation for Zebra 3, which is still very much a factor in my life.

I’ll be honest. I wasn’t blown away with Hive when I first tested the beta version three years ago. That being said, I wasn’t the same synthesist three years ago that I am today. Plus, I had spent so much time in Zebra that I had developed a nasty habit of holding other instruments to its unrealistically high standard of flexibility. Now that I’ve had a chance to spend a little more time with Hive, it has actually become one of my all-time favorite synths for reasons I couldn’t appreciate three years ago.

The Review

The first thing that flew over my head: the single-page interface, which wasn’t so obvious at first because it’s not really something you need to look for… and that’s pretty much the whole point! Everything you need (with exception of the Mod Matrix and the Arp/Sequencer and Effects pages in the Hexagon display) has been laid out in a gorgeous single-page user interface that allows for the same kind of unencumbered workflow that people enjoy when working with hardware analog synths.

Another thing you might overlook is that Hive has three synthesis engines (Normal, Dirty and Clean) that affect Hive’s architecture in ways that totally change its sonic character. In Normal mode, oscillator detune becomes exponential, envelopes have “s-shaped” attack and short decay, and the filters are oversampled, self-oscillating “ladder” models with nonlinear resonance. In Dirty mode, oscillator detune becomes evenly spaced, envelope stages are exponential, and filters are oversampled, self-oscillating “diode ring” models with very unpredictable and highly nonlinear resonance. In Clean mode, oscillator detune is slightly wider, envelopes have linear attack and exponential decay and release, and the filters are linear (non-distorting) “state variable” models with no oversampling. So yeah, there’s actually a lot going on behind the scenes.

Obviously, the first thing that jumps out at you is that unison has been handled in a very different way than the “stacked” voicings (single, dual, quad, eleven) within Zebra. Now, you can select anywhere between sixteen unison voices per oscillator, and as I explained earlier in my description of the three synthesis engines, the spacing between unison voices can be random, uniform or exponential.

The oscillators themselves are pretty simple, as they were meant to be, with seven basic waveforms: sine, saw, triangle, pulse (with a hidden Pulse Width parameter in the Mod Matrix) square, half, narrow, and two standard variants of noise: white (bright) and pink (dark). Also, there are hidden Oscillator Tune parameters in the Mod Matrix, even though there are discrete tuning controls for Octaves and Semitones in the Master Oscillator section, with additional controls for Vibrato (linked to an independent Vibrato LFO) Detune, Volume, Pan, Width for unison voices, and also Volume and Semitone controls for a dedicated Sub Oscillator, which has exactly the same waveforms and noise sounds I catalogued earlier, not including “like Osc”, which will inherit the waveform and unison settings in the main oscillator.

If you set both the main oscillator and the sub oscillator at the same pitch, they will have a 90-degree phase offset, regardless of the current phase setting (reset, random, flow) with exception of “like Osc”, which is synced to the phase of the main oscillator. Also, there’s a hidden Oscillator Phase target in the Modulation Matrix that you can use to offset the phase with a “Constant” source, or modulate the phase offset with an envelope or LFO to create phaser effects, which are especially useful when programming vintage bass and pad sounds. Speaking of all things vintage, I really appreciate the “Flow” phase mode, wherein the phase of a note picks up where the previous one left off, which is somewhat similar to the phase relationship between oscillators in classic analog polyphonic synthesizers.

The filters are made to be low on CPU, with bipolar Input Gain controls that determine the input volume and also the amount of nonlinear distortion when the synth engine is set to Normal or Dirty mode. There are seven different filter modes: Bypass, Low-pass 24, Low-pass 12, Band-pass, High-pass, Band-reject, and Peaking, which is similar to Bypass mode but with active resonance. A row of buttons are input selectors for any one of the oscillators/sub oscillators, and an extra input for a serial routing configuration between Filter 1 and Filter 2. There are also pre-routed filter cutoff modulation controls for both LFOs and both Mod Envelopes, and obviously the basic controls you can expect to see for Cutoff, Resonance etc.

The waveforms in both LFOs are much like the waveforms in both oscillators but with up/down saws and low/high square waves, a random hold (similar to a sample & hold signal) and a random glide waveform. Beside the waveform menu, there is a “+” button that, when set to “positive”, will output a unipolar LFO signal. In the restart menu, there are four restart modes: “Sync” mode will synchronize all voices to the host, therefore all voices snap to the same phase value, but you can modulate the phases apart with a polyphonic control signal like KeyFollow or MIDI note velocity. “Gate” mode will restart the LFO for each voice in accordance with the Phase knob, whereas “Single” mode will restart the LFO with the next MIDI note after all previous notes are released. “Random” mode will ignore the Phase knob, restarting the LFO for each voice at a completely arbitrary phase value.

Seriously, the attention to detail in regard to oscillator phase is nothing short of astonishing.

As I mentioned earlier, Hive’s envelopes behave differently depending on which of the three synthesis engines are in use. Even though I covered this before, it bears repeating: In Normal mode, envelopes have “s-shaped” attack and short decay. In Dirty mode, envelope stages are exponential, and in Clean mode, envelopes have linear attack and exponential decay and release. On either side of each envelope are velocity sensitivity controls, and in the trigger menu, there are options for MIDI gate, both LFOs and “One Shot”, which will cycle through all envelope phases, ignoring the gate length of each played note.

The Arpeggiator / Sequencer (next to the Effects tab within the Hexagon display) can be used in ways that we haven’t seen before in u-he’s previous synths, including the ability to record live MIDI notes and velocities into the sequencer, a Modulator mode that uses CC (continuous controller) values attenuating live MIDI control signals in time with the sequence, and also gate percentage and envelope parameters scaling the “SeqGate” source available in the Modulation Matrix. It just might be the most sophisticated sequencer I’ve used so far… Okay, perhaps with exception of Bazille’s modulation sequencer, but you can’t record live MIDI into that one, plus it’s not nearly as easy and fun to use.

I do believe that Hive was the very first u-he synth that implemented drag and drop modulation assignments (similar to Native Instruments Massive) via the target fields in the Modulation Matrix, which has been split up into two separate pages with six slots, each with two targets per slot. Also, there is a “Via” secondary modulation source that, if selected, will attenuate the primary modulation “Source” passed on to the two available modulation targets, each with a dedicated “Depth” knob.

One thing that I love about Hive is how quickly you can program modulation for effect parameters, and how seamless and bug-free the results are. I’ve noticed that effect parameter modulation within Zebra 2 can produce undesirable artifacts, especially when modulating reverb and delay parameters, so I’m very happy that these artifacts are not present within Hive.

In the Effects section, we have a Distortion module with four modes: Soft Clip, Hard Clip, Foldback and a “Corrosion” sample rate reduction (bitcrusher) mode. The Chorus module is notably different than the one provided in Zebra, with three modes: Dramatic, Classic and Ensemble, which is especially rich and swirly. The Reverb is significantly more modern than the reverb in Zebra, which might inherit Hive’s plate reverb algorithm once Zebra is updated to version 2.8 (hopefully) very soon! The Phaser module is somewhat similar to the one provided in Bazille, with the same “Stoned” and “Flanged” modes we know and love. The Delay module is quite flexible, with host-syncable delay times and the same basic controls we’ve all seen before in u-he’s other instruments, except this time we have “Width” and “Diffusion” parameters and a “Pong-Ping” mode which is basically the polar opposite of Ping-Pong mode. Also, there’s a 3-band EQ with low/high shelving filters and a fixed-width mid band, all of which have frequency and gain controls, and last but certainly not least, a powerful Compressor module with a small handful of intuitive controls and a gain reduction meter that helps you visualize signal attenuation. You can rearrange any of these effects in any order, a popular feature inherited from Bazille.

The Verdict

Hive might be the single most workflow-efficient, high-performance synth I’ve used with this amount of flexibility and pristine sound quality. You can dive right into sound design without fidgeting with annoying hidden pages and distracting context menus, and the sound! There’s a delicate balance of analog warmth and digital dirt, with crisp, fizzy highs and a fat low end that really punches you in the chest. There’s just no other way to put it… I absolutely adore this synth!

More info: U-He Hive ($149)