25 software synthesizers and sample players in one tidy package.
If you've been curious about Native Instruments' Reaktor synthesis and sampling software but you're reluctant to take the plunge into its complexity, Dynamo might be a good place to start. A collection of 25 software synthesizers and sample players created with Reaktor, Dynamo doesn't let you build your own devices from scratch, but does allow you to create and save banks of presets for each Dynamo instrument.
The 25 Dynamo virtual devices are taken from Reaktor's highly regarded Premium Library, and the upgrade path to Reaktor is reasonably priced should you venture into deeper waters. (For more on Reaktor, see "Master Class: Building a Reaktor" in the September 2000 issue of EM.) The Dynamo virtual instruments fall into six categories: drum machines (3), audio-processing effects (1), sample players and manipulators (7), classic/subtractive synths (5), FM and hybrid synths (5), and synths with built-in sequencers (4). We'll take a look at each group, but first let's see how Dynamo works.
INS AND OUTSDynamo can be used as a stand-alone program or as a VST 2.0 plug-in (if your digital audio sequencer supports that format). In addition to its own audio drivers, Dynamo supports ASIO on Mac and PC as well as MAS and DirectConnect on Mac. When running by itself, MIDI input and output is handled by OMS or FreeMIDI on Mac and by any installed MME MIDI driver under Windows. MIDI to and from a sequencer running on the same computer is also handled by OMS or FreeMIDI on Mac and by Dynamo's own MIDI driver (included) for Windows. I used a beige Mac G3/300 with 384 MB of RAM running under OS 8.6; relevant peripherals included an Emagic AW8 sound card and Mark of the Unicorn's MIDI Express XT MIDI interface.
Dynamo features a built-in, rudimentary MIDI-file player and an audio "tape deck" for direct recording to RAM (and subsequent saving to disk). Playing a MIDI file within Dynamo and then recording the results yields sample-accurate timing. Using the internal MIDI-file player gives lower latency and more voices than if Dynamo and a sequencer ran simultaneously. On the downside, the player provides only Start and Stop buttons, and when you double-click Stop it jumps to the beginning of the file. (A position display as well as Fast Forward and Rewind buttons are documented in the manual and slated for future release.)
Dynamo's audio recorder can play back recorded material and can also play audio files from your hard drive. However, this feature has limited utility because prerecorded audio files and Dynamo can't play at the same time, unless you have the right hardware and software. (Certain PC sound cards, such as the E-mu APS, can play two ASIO audio streams simultaneously. On the Mac, you could use a DirectConnect driver and an ASIO driver to get the same results. Nevertheless, with so many possible combinations of hardware and software, there's no guarantee this will work on your computer.)
As with Reaktor, the devices in Dynamo's collection are called Ensembles; they are able to contain a single sound-generating source called an Instrument, or they can contain multiple Instruments. Individual presets or programs are called Snapshots, which you can create on the Instrument and Ensemble levels. (An Ensemble Snapshot is just a collection of Instrument Snapshots.) An extensive collection of factory Snapshots is available for each Ensemble, and playing through them is an excellent way to get to know an Ensemble's sound.
Snapshots are saved and recalled from Dynamo's Toolbar. In Fig. 1, the name of the Ensemble Snapshot, HP Stringy, appears at the top of the screen. The selected Instrument is SalEM2 (marked by a red border around its name), and the Snapshot used is called SynthBass3. Dynamo offers an extensive file-management system for Snapshots; you can save, load, and merge any number of Snapshots for an Instrument or for the Ensemble. Once the Snapshots are loaded, you can recall them using MIDI Program Change messages.
At the top left of the Toolbar is a button that stops all audio processing; you can also do this by pressing the 0 key on your computer's numeric keypad. To the right of this button is a CPU usage indicator. Dynamo wants a lot of your CPU, even on a fairly fast system. Naturally, it gives top priority to audio processing, while relegating other tasks - such as scanning the computer keyboard and mouse - to the back burner. I was locked out of my system several times while playing Dynamo's demo MIDI files. You can lighten the processing load by reducing the internal sampling rate; this is recommended if you can live with less-than-CD quality. (Common rates from 22.05 to 132.30 kHz are supported.) Regardless of the internal rate you choose, Dynamo converts audio to your sound card's sampling rate on input and output.
Another way to save CPU cycles is to turn off some of the Instruments in an Ensemble. To maximize CPU efficiency, Dynamo recompiles an Ensemble each time a processing element is added or removed from the signal path. The drawback to this approach is that changing Snapshots can be a slow process when recompiling. This can result in a hiccup when playing back MIDI files that include Program Changes or when you manually change Snapshots.
The onscreen controls respond to vertical mouse motion. Under Windows (and slated for a forthcoming Mac update), the selected control can also be manipulated from the computer keyboard. This is a handy feature, because fine-tuning with the mouse can be tricky. You can also assign any of the controls to a user-selectable MIDI controller message - which means that you can automate parameters from a MIDI sequencer or hardware controller. Most of the controls in Dynamo's Ensembles have default assignments, but double-clicking on the desired control can change them.
You can also click the MIDI Learn button (a MIDI-plug icon with the letter L) in the Toolbar to assign the next incoming MIDI controller message to the selected onscreen control. The feature is extremely useful when tweaking patches, because making temporary fader assignments in the software means you don't have to modify the controller mappings on MIDI hardware. To keep the new assignments, save the Ensemble under the same name or a different one.
Onscreen help can be toggled on or off with a Toolbar button; when help is activated, mousing over a control brings up a description of its function. Because the manual's explanation of the Premium Library Ensembles is somewhat cursory, this onscreen help is crucial to finding your way around some of these monster synths.
CAST OF CHARACTERSAs I mentioned earlier, Dynamo's Ensembles are organized into six categories. Some, such as the Ensembles in the Synths - Subtractive category, are modeled after vintage hardware synthesizers, while others, including those in Samplers + Transformers, reflect the unique features of Reaktor. Here is a summary of what's on hand.
Synths: Subtractive. There are five Ensembles in this category. The Ensemble called 3-oSC is a generic 3-oscillator synth with a multimode filter and ADSR envelopes for the filter and amplifier. On my Mac G3/300, 3-oSC was easily able to generate eight voices at a sampling rate of 44.1 kHz with plenty of CPU cycles to spare for my sequencer. In spite of - or perhaps because of - its simplicity, 3-oSC can generate lots of nice vintage analog sounds.
At the other end of the CPU-hog spectrum is Uranus, a 3-oscillator, dual-filter Ensemble with complex modulation routing and a number of built-in effects, including a beautiful 4-delay chorus. Uranus defaults to three voices at 33.075 kHz, but its lush pads are worth the price.
The remaining three synths are modeled after vintage analog machines: the Minimoog (ManyMood), the Oberheim 2-voice (Me2SalEM), and the Roland SH-101 (SH-2k). These software versions are not emulations of those synthesizers' exact sounds, but Native Instruments has attempted to model their idiosyncrasies.
My favorite among these three is Me2SalEM (probably because the 2-voice was my first synth); its components are shown in Fig. 1. The best-kept secret of the 2-voice was that lurking behind each of the two synthesizer-expander modules (SEMs) that produced its two voices were little plastic Molex connectors for patching the various components together. Me2SalEM brings a similar spirit to the front panel, with an input matrix that routes noise, an external signal, or the output of either SEM to the input of the other. A stereo delay has been added at the end of the signal chain for good measure. All three models can produce four voices without taxing the CPU.
Synths: FM + Hybrid. Three of the Ensembles in this group offer variations on an FM theme; another is modular with matrix patching; and the fifth is modeled after the PPG and Waldorf wavetable synths.
Of the three FM synths, FritzFM is the closest to a classic model. It has six sine-wave operators and produces most of the typical FM sounds. InHumanLogic also has six operators, two of which offer saw and pulse waveforms. It provides a switching matrix for various modulation algorithms, including complex output configurations. Although it is capable of traditional FM sounds, it stretches far beyond these limits.
Cube-X's four operators feature all the standard oscillator waveforms, along with noise and sample playback, and in this way it deviates most from the classic FM model. A 4-by-4 switching matrix allows any operator to modulate any other (as well as itself). Each of these hybrid FM models has a battery of effects; common among these are forms of distortion, chorus, reverb, delay, phasing, and multimode filtering.
As its name implies, MatrixModular is modeled after classic modular synths such as the EMS Synthi A, whose patching is controlled by a matrix rather than patch cords. MatrixModular's sound sources include two multiple-waveform oscillators (which can be synchronized), a noise generator, a sample player, and a granular sample-resynthesis module. External audio can also be processed. The control modules include ADSR and AD envelope generators and a multiwaveform LFO. There are two 4-pole filters, a ring modulator, and distortion and delay effects for signal processing. You also get a versatile 16-step sequencer with three control/pitch tracks and separate gate tracks for two envelopes, sample trigger, and LFO reset.
The heart of MatrixModular is a 16-by-16 modulation matrix that lets you route anything to anything else and is easy to operate. As you might imagine, a synth with a preset named Insekt Jazz is capable of some pretty bizarre sounds. The CPU load depends on which modules are in use, so you can get some very economical patches out of MatrixModular. If it weren't for deadlines, I'd still be playing with this synth.
The last synth in this group, Nano-Wave, emulates the PPG and Waldorf wavetable synthesizers. It comes with 43 wavetables that are available to two sample-player modules. The key to wavetable synthesis is controlling and modulating the sample-location pointer in the wavetable. In NanoWave, this can be controlled by any combination of ADSR envelope, LFO, key number, and Velocity. The frequency of each sample player can be modulated by the other, and a noise generator and sine-wave oscillator provide additional sound sources. Output processing includes a multimode filter, ring modulation, chorus, and delay. Due to the variety of the included wavetables, the sample players alone give you an interesting array of sounds, and NanoWave is quite CPU efficient provided the other signal processors are turned off.
Synths: Sequenced. Each of the four Ensembles in this group has one or more built-in 16-step sequencers. These Ensembles are distinguished by their sound generators and sequencer routing. NewsCool and SineBeats are classic beatboxes. NewsCool has four sine-wave tone generators, and SineBeats has three as well as a noise source. NewsCool's sequencer is especially simple, with a single knob for each stage to control the triggering of any combination of tone generators. SineBeats has a separate 2-track sequencer for each tone generator: one track controls Velocity, and the other track controls pitch and decay-time modulation.
Cyclane combines a NewsCool-style beatbox with a separately sequenced FM synth for pitched sequences. The FM sequencer has a scale-correction section that conforms melodic patterns to a user-specified scale. The sequencers can run forward, backward, or bidirectionally; individual steps can be gated on or off; and even-numbered steps can be "grooved."
NewPrimitive departs the most from the beatbox model, with one 2-track sequencer driving a generic subtractive synth and another driving a sample player. The first track of each sequencer controls Velocity, and the second controls modulation amount. Modulation routings include filter cutoff, pitch, and two distortion parameters for each synth. For the sample player, sample start is also a modulation destination. NewPrimitive is the most economical of the four Ensembles in this group, and it also produces the most radical (read: grungy) sounds.
Drum Machines. The drum machine called Gonzzo maps its eight sample players to the white keys from C4 to C5. It comes with a sample set that contains 79 drum sounds (you can also import your own samples). Gonzzo's output processing features reverb/delay, 3-band EQ, resonant filtering, and compression.
Drumatik and DSQ-32 use oscillators designed for individual drum sounds. Drumatik uses the General MIDI note map for its eight oscillators, and DSQ-32 uses the same map for its seven. DSQ-32 also includes a built-in 7-track, 32-step sequencer. Fig. 3 shows example tone generators from all three drum machines.
Sample Players. This group contains seven unique sample players (the company refers to them as samplers). Beat Breaker and 6-Pack use integrated sequencers and are designed primarily for loop playback and manipulation. The 6-Pack sample player has four loop players and two sequenced sample players. With its straightforward design, synchronizing the beat loops and sample sequences is easy. Beat Breaker has a loop player and a granular sample-resynthesis module. Its 4-track sequencer controls loop selection, position, transposition, and volume. It also features a complex "sound shaping" section that will quickly take you to the crumbling edge of techno.
Formantor, Plasma, and Triptonizer each take a different approach to granular resynthesis. Formantor gives you independent control of pitch, speed, and formant shifting. Plasma provides independent control of grain position and speed, and it is well suited to creating lush pads from almost any source material. Triptonizer has separate envelope generators and LFOs for position and formant shift during sample playback; it also includes a complex stereo delay processor.
The remaining two sample players, rAmpler and Random Step Shifter, bring Dynamo back to earth. rAmpler is a straightforward sample player - although it does have a granular resynthesis module. Random Step Shifter is the simplest Ensemble in the kit: it breaks a sample (typically a beat loop) into quarter-, 8th-, or 16th-note segments, then randomly rearranges their playback order.
Effects. There's only one effects Ensemble: GeekFX. It has a sample player to load samples for processing, but it can also process external audio input on the fly. The effects are divided into three sections - distortion, filter, and delay. Each offers an interesting selection, but the distortion effects are the most unusual.
FINAL TAKEDynamo offers a lot of value for your software-synth dollar, but you'll need a fast computer. And even with a powerful machine, the program doesn't leave much room for other processing (say, by a digital audio sequencer). I found Dynamo easiest to use as a stand-alone application for creating audio files to be used later in a sequencer. I was able to use Dynamo successfully as a VST plug-in, but its CPU demands limited the utility in that context. I also found it to be very playable (that is, with low latency) on both laptop and desktop G3/300s using the Mac's built-in Sound Manager drivers.
The 25 Ensembles that come with Dynamo provide a diverse palette of sound- and sequence-design tools. Exploring them is educational and a lot of fun. The Dynamo manual takes a decidedly minimalist approach in describing the Ensembles, but the onscreen help picks up the slack - think of it as on-the-job training. Considering that you get a kit of soft synths ranging from classic virtual-analog to granular synthesis, something in Dynamo is bound to get your attention.