Review – The Drop from Cytomic
If you are looking for a dedicated filter plug-in, then our reviewer suggests you Drop whatever you’re doing and check out Cytomic’s superb offering.
by Per Lichtman, May 2017
Cytomic The Drop ($99 USD, download from Cytomic.com) is a resonant high-pass and low-pass filter FX plug-in emulating in extensive detail the filter behavior of ten different filters, many taken from well-known synthesizers (and others from less common designs), with the ability to switch between higher CPU and lower CPU models of each. Both PC and Mac are supported and all standard mainstream formats are available.
In some ways this is one of the simplest reviews I’ve ever written, because after spending time comparing dozens of different resonant filter plug-ins, The Drop is simply the most accurate dedicated synth FX plug-in I’ve ever tried and it works as a great saturation plug-in, too. On top of that, there’s a time-limited demo available so you can try it yourself right off the bat. So if you ever wanted to get as close to the sound of some of these filters as you can get in-the-box, this is pretty much the place to go. At the same time, it’s one of the more difficult reviews I’ve written because this filter plug-in can do so many different things that even the 87 presets included only give a slice of what the plug-in can do. In other words, it’s easy to use but you can use it for much more than you would expect …
For Those That Want a Quick Sonic Tour…
Since The Drop is capable of both bread and butter sounds and some really complex ones, I want to mention that there’s a yellow drop-down menu at the very top that says “Default.” If you click on it and hover over the word “Presets” near the top, you’ll see eight categories of presets with a total of 87 presets ready to load. The preset categories are Basic; Drum; Guitar; Keys; MIDI; Mix; SFX and Synth. These presets help quickly show some of the diverse things the plug-in can be used for – though the list is far from comprehensive. From using the plug-in to add some high-end sheen to a mix (the oversampling options make it surprisingly good at boosting the high with resonance) to various auto-filter style sounds and complex modulations, you can take your sound (or whole mix) in a lot of different directions just by cycling through the presets. Of course if you want to start turning the knobs or tweaking the settings yourself (and who doesn’t) then you may want to know what the available filters are based on.
Filter Models and Saturation Options
As of version 1.5.2 (the most recent version at the time of writing) the ten unique circuit names are mostly taken from the vintage synths emulated: MS2 (Korg MS20 rev2); SMP (custom halfway between Korg MS20 rev2 and OSCar); OSR (OSCar); MS1 (Korg MS20 rev1); JPR (Roland Jupiter 8 or Juno 6); SHR (Roland SH2 or SH101) and PRD (Moog Prodigy). You may have noticed that, that list only contains seven models, but three more were added at the end of last year: AMU (a custom design “inspired by the ARP Odyssey Mk I filter”); KSM (“a custom design input mixing Sallen Key”); WSP (from the EDP Wasp “with unique CMOS inverters”). The AMU model is one of my personal favorites.
Each filter has a different sound, with the differences ranging from subtle to massive depending on the specific filters, the settings and the source material. The differences are especially pronounced when using the filters at high resonance settings that cause self-oscillation.
The full CPU usage models are engaged whenever the “HD” button is highlighted in the LP or HP tab, but by pressing the button again you can bypass HD mode to engage a more economical model. It’s also possible to use the economical mode during editing but select an option in the oversampling menu to automatically switch to HD mode during render. In my testing, I found the more economical model tended to reduce CPU usage by somewhere in the 15-25% range.
Each circuit is modeled with 12 dB/octave and 24 dB/octave slopes for both high-pass and low-pass filter. The LP and HP filter can be enabled and disabled independently and there is a drop-down menu to switch between the default serial mode (where the output of the HP filter is sent to the LP filter) and parallel (where the two filters operate on the signal simultaneously). Every model is capable of in self-oscillation in both filters and it’s actually even possible to use the self-oscillation for MIDI controlled pitch generation (more on that later) but there is a “safe” switch that can be engaged to keep the resonance knobs from activating self-oscillation. The self-oscillation range on the resonance knobs is indicated with an orange highlight as you pass 100%. Next to the resonance knob is a small gray arrow (small enough that you might miss it) that acts at the trim pot for the self-oscillation resonance range. When it’s turned all the way to the left, the orange self-oscillation range spans roughly 101-125%, but you can turn the trim pot all the way to the right to make the range span 101-200% for some even more extreme resonance.
The filters differ noticeably in the curves of the filters themselves, which is shown in the visualizer display and which can be auditioned by swapping models in the drop-down menu while keeping the settings the same. There are also HP drive and LP drive knobs that let you increase the amount of saturation that the filter circuit provides for the signal. The way those knobs work is by increasing the input level by the same amount as they decrease the output level for the filter, allowing the level to stay pretty similar while increasing the saturation.
Level and distortion characteristics vary greatly between the filter models, even when a filter is set to be as transparent as possible (for instance by setting the low-pass cutoff to maximum and resonance to minimum). The OSR and SMP circuits have a higher ceiling on their level than the other models, with the SHR and PRD circuits having the lowest ones. In my personal experiments, I found that there was a difference of up to 11dB in the output level when I massively overdrove the signal through the highest and lowest circuits. But the ceiling is only part of the difference – the character also varies greatly as well.
Elegant and Customizable Design
There are so many things about the way that The Drop is designed that you’ll find you appreciate more as you work with it. At a basic level, you’ll find the usual niceties like A/B comparison for two sets of settings, responsive controls and sufficient contrast, and a visualizer. But there’s also the extensive control for setting different quality settings (the level of oversampling; whether HD mode is on or not) both while working and for the final render. The oversampling goes from x1 (which is off, meaning the filter runs at 44.1/48 KHz depending on your session) up to x16 (705.6 kHz) in real-time modes, or up to x64 (2.8 MHz) for off-line rendering – so you can completely avoid aliasing artifacts unless you want them. You can also use different quality settings in each plug-in instance or copy your current settings to all the instances in your session – or save your current quality settings as the default for any new instances. Then there’s the way you can custom scale the GUI to different resolutions by dragging the lower right corner (or just selecting sizes from 50% to 200% from a drop-down menu), making it easy to scale up the size to more easily read some of the smaller labels, or scale down to make the interface take up less space. You can turn “show CPU usage” on or off, and will show how much CPU each of the two individual filters is currently using. You can easily see the folder where your presets are stored by clicking “show current folder” so you don’t have to dive for the manual to figure it out, making it easy to back presets up or share them.
Comparing Filter Saturation
An unexpected bonus in Cytomic The Drop is that the differences in gain response between the models are actually pronounced enough that I found myself frequently loading the plug-in just to use it for the saturation colors alone. To that end, I wanted to look at the colors the models presented.
Back in 2016, I did some tests with a loud MicroTonic drum sequence to emphasize these gain emulation differences between the seven models available in The Drop at the time, without over-emphasizing the differences in the filter curves themselves. To do this, I put the filters in serial mode, set the HP and LP drive knobs to maximum, set the HP cutoff to minimum, LP cutoff to max and both the resonance knobs to 0% – then I would switch the LP and HP models to the same circuit at once. Then I manually brought the output level up, using the post-knob, to bring the peak dbFS up to the within half a decibel of the original dry control signal (never exceeding it). In all cases, the signal was being pushed hard enough that HP and LP drive LEDs were red at least part of the time, but not so hard that they stayed red the whole time.
Comparing the MS1 to the dry control signal, the sound was much “punchier”, with the bass drum sitting much more forward in the mix, a narrower dynamic range and somewhat darker sound overall (albeit with some addition the high end from the distortion harmonics). Using the same method, the MS2 seemed to be saturating less at the same level (which the LEDs reinforced), preserving more of the original dynamic range and frequency profile compared to the MS1, but still coming with a “louder”, “fuller” sound that had a narrower dynamic range than the dry signal. The SMP unit yielded a still darker sound than the MS2, being the smoothest so far and quite “rounded” compared to the dry signal – but again, with far more dynamic range than the MS1. The OSR sounded darker, yet again and the MS2, SMP and OSR circuits all sounded far more similar to each other than any of them to the MS1. The SMP and OSR options both took some of the edge off of the hi-hats more than the MS2. The JPR HD model provided much stronger saturation, more in line with the MS1, but noticeably darker and rounder – with either model providing a massive increase in subjective loudness at the same peak dbFS as the original dry signal. The SHR provided the most strongly colored sound yet – very up-front and bringing out the sustains of the drums, de-emphasizing the transients in comparisons to the other options so far. Arriving at the PRD, we find an interesting difference from SHR, though both are strongly colored: whereas the SHR was saturating the HP filter more than the LP filter with the test settings, the PRD was saturating the LP filter more than the HP filter. Perhaps that is why the PRD yielded a brighter sound than the SHR, though it was similarly strongly colored.
Looking back at the results of the experiment, we see can see that the models can broadly be grouped into two camps in regards to their saturation use. There are the high threshold models that color the signal less at lower input levels (MS2, SMP and OSR) and the low threshold models that start to saturate more dramatically at lower levels (MS1, JPR, SHR and PRD). Out of the low threshold models, the MS1 was brighter than the others, with a wider dynamic range. The low threshold models were great for things where I wanted a heavy color (like the synthetic drum loop used for testing) but for more melodic material, I found I gravitated to the MS2, SMP and OSR. In both cases, the “Pad” button in the controls provides a quick way to change the amount of saturation in 6dB increments from +12dB to -12dB, making it easy to adapt to differences in the thresholds if you change your mind about the filter model you want to use.
Massive Modulation Options
Despite the deceptively simple GUI, I fear I can’t fully do justice to the number of modulation options on offer. Sure, you could use automation to sweep the cutoff and resonance knobs and other basics like that or use an LFO or envelope in a basic way to modify the frequencies, resonance amounts, pre-gain or post-gain but almost without exception, you’ll find that you can dig much deeper than you expect. LFOs offer tempo syncing, rate, phase, spread and a multiplier control: the multiplier control lets you either slow the LFO down or speed-up to 32,768 times the base LFO rate, whether tempo-sync is engaged or not. LFO has a plethora of additional controls, like swing amount and curve adjustment – and surprisingly The Drop even has its own step-sequencer with from two to 32 steps and the option to switch between sixteen patterns. The step-sequencer is edited in the same blue window at the top-center of the GUI by dragging the steps up and down with the mouse.
Even without getting comprehensive about all the controls, there are several important highlights to cover, like the FM modulation (which can either use the normal audio or a side-chain, processed raw or via an envelope or LFO 2) or the MIDI tracking (which in combination with the self-oscillating filters makes it possible to turn the filter into a tonal instrument of its own). In fact, at one point I actually constructed an organ using several instances of the plug-in to create sine-wave style oscillators off the harmonic series. You can do some really exotic things! Needless to say, it would be impossible to catalog all of them.
A few years back, I went through and auditioned just about every FX filter plug-in I could my hands on. While I found several interesting ones (many of which I enjoyed using for particular types of material), none really sounded close to the analog filters I had spent time with “outside the box.” More recently, I found certain synth plug-ins (u-He Diva for example) did noticeably better than previous virtual instruments in regards to more accurately modeling analog filters. But even u-He’s Urs Heckmann made special note of the quality of work that Andrew Simper did with The Drop’s filters, and having spent time with The Drop, it’s simply become the first analog filter plug-in I reach for every single time. To be clear: that does not mean it does everything. So what might you look to another plug-in for?
One of the first things that comes to mind is those looking for a more “digital” and “less analog” sound, as there are other filters that offer types of AM and FM sounds other than those easily accessed in The Drop. Also, there are no dedicated formant or vowel filters. Similarly, The Drop does not offer a native bitcrusher or dedicated waveshaper, so those looking for multi-FX filters incorporating those can look elsewhere. Plug-ins I came across in my earlier search that addressed those areas included Tone2 BiFilter 2, SugarBytes Wow and PSP Nitro.
Another area is those specifically looking for extremely steep filter curves, for which I would point PC users to the free RubberFilter by Christian Budde (capable of curves as steep as 384 dB/octave). And for those looking to use more filters at once in a single interface (or just prefer a different GUI design) there’s FabFilter’s Volcano 2.
Each of these plug-ins has its own strengths, but no plug-in I’ve used captures the sound and behavior of the analog filters modeled in The Drop as well as The Drop itself. In my eyes (and to my ears) it represents the current state-of-the-art.
Is It Right For You?
Cytomic The Drop is, by far, the best dedicated plug-in for analog style resonant filters I’ve ever used – so if that’s what you’re looking for, it’s the very first one you should look at. With ten different models on offer for both high-pass and low-pass types in 12dB/octave and 24dB/octave versions, you can quickly cycle through different variants of your filter sound. It’s the best emulation of the filters modeled I’ve ever heard in a plug-in. The modulation abilities (step-sequencer; AM style effects via the LFO; envelopes; MIDI note tracking) are also great and there’s even FM on offer. CPU usage can easily be modified by switching between full and reduced CPU models, or by modifying the amount of oversampling. What’s not to like? Basically, if you’re looking at filter FX plug-ins you owe to yourself to check out The Drop.