Review – Diversion by Dmitry Sches
Diversion is a software synth that offers subtractive capabilities as well as some elements of FM, granular and RM synthesis. Learn more about it here.
by Rob Mitchell, Nov. 2014
Diversion is a softsynth plugin by Dmitry Sches. In its basic form, it is a subtractive synth, but has some elements of FM (frequency modulation), granular synthesis, and RM (ring modulation) as well.
It was first released in August of 2011, and the plugin received a lot of praise on many of the online forums. It wasn’t just another virtual analog type of synth, which has been done a million times over. It had a special character to its sound, as well as a great user interface, and many people took to it right away.
It started out as a 32-bit Windows plugin, but soon after that, a 64-bit version was available. Eventually, Dmitry also developed a version for Mac users. He added sample support and granular synthesis in a new release for December 2013, adding to its impressive feature list. I’ll get to those features soon, but first I’d like to go over the installation and system requirements.
The installation was very easy, and you just place the personal license key into the folder where Diversion is installed. For you PC users, one great thing about the plugin is that it won’t mess with your registry at all.
The system requirements are on the high side, as they recommend a “modern powerful processor”. It is available in 32-bit and 64-bit formats for both PCs and Macs. For PCs, it will run using XP, Vista, or Windows 7. For the Mac, you’ll need OSX 10.6.8 (or later). As long as you have an Intel i5 or i7 CPU, then you should be ok.
They do mention that most of the presets were designed to be playable using an Intel Core 2 Duo P8600 processor. That is pushing it though, as they also mention it will be around 50% CPU usage when playing most of the presets in Diversion’s library. You could always freeze tracks in your DAW, and it may be a good idea, especially if you’re going to use many instances of the synth plugin.
One reason they mention the higher specs is the switchable oversampling. If you switch it to one of the higher oversampling rates, such as 4X or 8X, it sounds great, but it will also use a lot more of the CPU. The standard way it is set is 2X, so many presets should work ok on slower PCs or Macs, but a faster CPU doesn’t hurt either. It also has up to an eight-voice unison, which you can adjust if things get a little out of hand. Some useful tips on how to reduce CPU usage are given in the manual.
Diversion is priced at $169 USD and includes nearly 400 presets. You can download the demo version from their website to try it out with your own setup. The only limitation on the demo is that it plays a noise every 90 seconds.
Diversion has four oscillators, and unlike many synths out there, the main waveforms are generated by the oscillators themselves. Each oscillator has a power button, so you can switch on just the ones you want to use. This can help when you’re creating your own presets, as you can easily isolate one oscillator’s sound.
Some of the oscillator controls are standard ones you’ll see in other synth plugins, such as tuning by octave, semitone, or percent. In addition, there is a phase control, a “free” switch to change the oscillator to free-running mode, and a switch to invert the polarity for the waveforms as well. The Vibrato control adjusts the pitch modulation amount from the LFO section. I’ll mention more on the LFOs later.
Each of the oscillators has its own X/Y controller. Depending on the waveform selected, movement in the X direction will brighten the sound, and changing the Y direction will manipulate the tone.
There are many waveform types to choose from, and as I mentioned before, their audio is created on-the-fly, with a few exceptions that I’ll get to in a minute. When you click on the waveform name, a menu appears letting you select between the built-in waveforms, samples, or wavetables. Of course, the samples and wavetables are not generated like the built-in waveforms. The only built-in waveforms that are not created from scratch like the others are of the noise variety.
The oscillator section can also apply some effects to the signal. In total, there are seven different effects, and there are four controls. Each of the four controls can be switched between the seven available effects, just click on its name below the knob to select one.
Some of the included effects are feedback, a sub-osc, a boost for the high frequencies, and Chip, which is a Lo-Fi, aliasing type of effect. Just for fun, I switched them all to Chip, set them to different amounts, and then modulated three of them using an LFO on each. It was quick and easy to setup, as you can just right-click on the control, and use the “Add Modulation” menu. Then it’s just a matter of going to the Modulation Matrix and turning up the amount to whatever level you’d like.
Speaking of the Modulation Matrix, there are 24 slots available, and as you change the level on one of those slots, the control that it is affecting will have a colored ring around it. For instance, if you turn the level halfway up (or down), that colored ring will go halfway around the control. This is a great visual feedback, and it helps to keep tabs on what’s going on during the preset design process.
Ok, now let’s get back to the oscillators. If you load in a sample, a few things will change on the display. Those same effect controls I mentioned earlier will switch to either being inactive when in sample mode (if switch is set to “samp”), or they will change into controls for granular synthesis (if switch is set to “grain”). When the granular synthesis is enabled, you’ll see these controls:
“Dens” – number of grains
“Tune” – variation of grain’s pitch
“Size” – changes size/length of every grain
“Pan” – pan amount for the grains
The first and third oscillators also have controls to add ring (RM) and frequency modulation (FM). The signal for the RM and FM are from the second and fourth oscillators.
In the upper right corner of Diversion, there is a record button. This allows you to record the output of Diversion, and then save it as a WAV file, or quickly send it to one of the four oscillators.
You also have access to a powerful, onboard wavetable editor. It has a total of eleven filters and distortion effects you can use to change up the sound to your liking.
Envelopes, LFOs and MSEGs
There are four identical envelopes you can use, each with the standard Attack, Decay, Sustain and Release controls. They each include a Curve control for the Decay and Release part of the envelope. This lets you use an adjustable exponential or variable-slope mode. The Speed control lets you change the maximum speed that the Attack, Decay, Sustain and Release controls can use. The AMP envelope is used to shape the amplitude of the audio, such as setting it to a slow attack and release for a pad sound. It can also be assigned in the modulation matrix to other targets.
Just like the envelope section, there are four LFOs available to use. Each has controls for amount, rate, phase, and speed. There are eleven different waveform settings, and depending on the Trigger setting, it can work in three different ways: “Poly” mode is a per-voice setting, “Free” mode is monophonic in nature, and every voice has the same waveform. The “First-key” mode is monophonic also, but if a non-legato note is played, it will be re-triggered.
Four MSEGs (multi-stage envelope generators) are on board as well, giving you customizable envelopes to use for modulation. 64 nodes can be added per envelope, and it can be looped. There are controls to zoom in or out, scroll left and right along the length of the MSEG display, and the whole envelope can be made to fit on screen by double-clicking the top bar.
Filters and Bus Processors
Next to each of the four oscillators is a multi-mode filter. There are tons of different filter types included. Some of these include an Analog modelled Low-pass, High-pass, Band-pass, Formant, Acid Low-pass (with its own saturation), Notch, Peak, Comb filter types, and more.
The filter section has some standard controls, such as Cutoff, Resonance, Key tracking, and Velocity. It also includes a Drive control, which can add a level of saturation. You’re not stuck with only one type of saturation though, as there’s a hyperbolic shaper, soft and hard clip, three types of asymmetric distortion, and a sinusoidal type.
At the bottom of the filter section for each oscillator is the output section. Here you can adjust the volume level, panning, and send the signal to either (or both) of the two bus processors.
There are two identical bus processors in Diversion, and each can manipulate the oscillator’s sound even further. It is nice to have the added options here, as it includes a stereo filter (same types of filters as the oscillator section), distortion, and Lo-Fi sections.
The filter in the two busses works differently than in the oscillator section. There are actually two mono filters that handle the incoming stereo signal. Using the Stereo control, you’re able to split the cutoff for the two filters. The stereo effect it can achieve through modulation is pretty cool, and there are many possibilities for tweaking here.
If you feel you’ve got a great sound already, you don’t have to process the sound through either of the busses at this point. You can just switch the power off for either bus.
In-between the two bus sections, there is a bus mode switch: “Poly” uses per-voice processing, and “Mono” which combines the voices before processing. Finally, there is a send control which is lets you adjust the output going from Bus1 to the input of Bus2.
Trance Gate and Arpeggiator
The Trance gate has a 16-step editor in which you can make up rhythmic patterns for your presets. You change the amount of gating that is added to the sound with the Mix control, or taper off the attack part of each step with the Smooth control. The Rate and Speed controls will change how fast the pattern plays, and the duration of each step. It is very easy to use, but if you’re not feeling too creative, they also include a decent number of gate presets you can load.
In the powerful Arpeggiator section, up to 32 steps can used, and there are an ample number of controls to manipulate the notes. The modes include Up, Down, Up-Down, Down-Up, Random, Poly, and Mono. You are able to transpose the notes up or down, set it to play up to a four-octave range, and change the duration and velocity of each note. If this wasn’t enough for you, how about being able to set a per-note gate time, and use a handy legato switch in-between each of the velocity sliders? They’ve also added a shuffle control, and a global gate setting to top it all off.
Diversion has ten effects you can load into its FX Matrix. The matrix has an adjustable amount of cells for the bus processors output. It can have up to seven effects for each of the two bus processors outputs, and still have one cell left over for an additional effect.
The standard way it is setup is with two rows of four cells each, and each cell can have an effect loaded into it. The output of those two rows then mixes together, and is fed into four more cells that can have effects loaded in as well. You could think of those last four as “master” effects.
The effects include reverb, delay, echoes (another type of delay with fewer controls), grain shifter, chorus, flanger, phaser, distortion, tremolo and an equalizer. You can load two of the same type of any one effect. Just for example, one row may have a reverb with a small-room sound in one of the cells. In the other row, you might have yet another reverb, but with a larger room size on it.
Even though Diversion has an impressive list of effects, one I’d still like to see added is a compressor.
As usual, I try to cover everything in a review, and end up adding in bits towards the end that are notable. I couldn’t think of too much else to add to Diversion, but here are a few things I’d like in a future version:
In the preset browser screen, it would be nice to have a “favorites” type of setup, or some other rating system.
For the granular synthesis, it would be beneficial if more options were added for the grains, such as speed and grain position.
I’d like to be able to use the X/Y section of each oscillator to optionally control other targets via the mod matrix.
The last request is the ability to save the settings of an oscillator, and then be able to load it into another.
Diversion definitely can get an edgier, sharper sound, and the main character of its audio output seems to lean in that direction. However, it can also pull off regular, old-style virtual analog types of presets as well. Just because it has a filter for each oscillator, and a bunch of different ways to distort/contort the sound, (plus the two bus processors for additional manipulation) that doesn’t mean you have to use all of its many assets at once.
The higher CPU usage in many of the modern synth plugins seems to be the new norm, and that’s the case with Diversion as well. Most users of this synth plugin will be dealing with that, but they have been warned by the statement on Dmitry Sches’s website: “WARNING! Diversion’s audio engine may consume lots of CPU power”. It also goes on to state the recommended PC or Mac setup, and there is the demo version that people can try before buying.
For me, I think it is worth that extra draw of power, and it has that special “something” to its sound missing in many other synth plugins. That cutting-edge type of sound also makes it a “love it or hate it” type of synth, as some people have mentioned on various musical instrument forums. I happen to love it, and really think you have to try it for yourself, and you’ll be treated to a modern synth plugin in action. As a matter of fact, after checking it out in detail, this is one of my favorite synth plugins to date. Try it out and hear the difference for yourself.