Eventide SplitEQ

An EQ that gives you separate control over the transient and sustain portions could be more useful than you think!

We’re spoiled for choice when it comes to EQ plug‑ins. Even those that come bundled with our DAW software work well and have extensive feature sets, or offer a more than adequate feature set. But Eventide’s new SplitEQ is no normal equaliser. Supporting the usual plug‑in formats for Mac and Windows and authorised via iLok (no physical dongle required), it seems to be underpinned by the transient‑separation technology Eventide first employed in their Physion plug‑in (see Paul White’s review in SOS June 2017, written before its name change: www.soundonsound.com/reviews/eventide-fission). And while the idea of separately EQ’ing the transient and sustain portions of a sound isn’t new in itself (I can’t be the only one to have used transient shapers and EQs in parallel) it’s the first time I’ve encountered an insert plug‑in that takes care of all this task in one place. That convenience makes it far more than a problem‑solver...

Signalling Intentions

Conceptually, SplitEQ starts with a Structural Split process: it divides the incoming signal into two separate processing paths, one of which operates on the Transients and the other on the sustain or, to use Eventide’s term, the Tonal part of the signal. Each path flows through an EQ stage, with six configurable bands plus dedicated high‑ and low‑pass filters. Each band also has an L‑R pan or M‑S stereo width control (though both can’t be used simultaneously for the same band), before the two paths are recombined to form the output signal.

Useful global controls including separate but linkable ±30dB level controls for the Transient and Tonal signals, a ±200‑percent EQ scaling facility (with separate linkable controls), a pan/width stage for each path, a global bypass and another bypass for the EQ only; engage the latter and SplitEQ can behave more like a conventional transient shaper.

The split process sounds transparent to my ears, in that when no EQ or stereo manipulation is applied and the global Transient and Tonal faders share the same setting, the signal at the output sounds the same as at the input. This splitting process is performed automatically but while you don’t need to tweak anything for it to do its thing, there are some useful controls for refining the split.

First, at the bottom of the GUI, you can choose different transient detection presets, each fine‑tuned to suit a different source (Full Mix, various drum/percussion options, Bass, Piano/Synth, Guitar, vocal and General). Helpfully, you can lock (or unlock) this parameter once selected: choose a different global preset and this selection will remain unchanged if locked.

A Transient Separation parameter determines the point after a transient is detected at which the source is split, and Transient Decay acts like a release control: longer times lengthen Transients and the attack/swell phase of the Tonal path. Finally, a Smoothing control can help you solve problems: lower values give greater differentiation, but if set too low, sound abrupt; higher values sound smoother but make the paths increasingly similar. Thus far, I’ve felt the need to tweak this only rarely, but it’s good to know the facility is there when needed.

Eventide recommend starting by choosing an appropriate source preset before moving on to these ‘refinement’ controls to target the specific source in question, and I’d agree that this is the best approach. To aid your decision making when selecting and refining these presets you can solo either the Tonal or the Transient signal, and this solo auditioning process is essential if you’re to get the best results from this plug‑in.

Band Controls

Once you’ve determined how the signal is best split, you can turn your attention to manipulating each path. As I mentioned in passing, there are eight bands in total: low‑ and high‑pass filters along with six more configurable bands, each of which can be set to act as one of six types (Low/High Shelf, Peak, Notch, Tilt Shelf, or Bandpass).

For each of the eight bands you can choose independently whether the EQ is applied only to the Transient signal, only to the Tonal element, or to both paths simultaneously. Obviously, operating on both from the outset means SplitEQ acts as a regular EQ, but note that if you’ve previously adjusted the Transient and/or Tonal paths, you can boost/cut both together, so that the difference between them is preserved. As with the global signal, you can solo the Transient or Tonal part in each band, to aid you in making these adjustments.

You have the option of making your tweaks by click‑dragging on, or typing in the parameter fields at the bottom of the GUI or, my general preference, click‑dragging nodes on the GUI’s EQ curve and using key modifiers to tweak certain parameters for the selected band. The way in which you select the nodes is pretty slick: hover over a band’s default grey/white node, and green and blue wings/tabs appear above and below it; click‑drag the grey part to manipulate the whole signal for that band, the green wing to change only the Transients, or the blue one for the Tonal path. (This colour‑coding is reinforced throughout the GUI, making it all very easy to digest). Once you’ve adjusted the Transient or Tonal component, you can click anywhere between them in the highlighted band to act on both simultaneously. Modifier keys allow you to adjust the Q (Shift‑drag), reset a node (Alt/Option‑click), or make finer adjustments (Ctrl‑drag).

There are some small limitations but nothing to really put me off, and some could be addressed easily in a future version. For instance, while you can set the Q and gain for the main six bands’ Transient and Tonal paths independently (I’ve often found that I want to boost one and cut the other, each with a different Q), those paths always share the same frequency. The high‑ and low‑pass filters can operate on different frequencies for each path, and extending this approach to the six configurable bands would make SplitEQ considerably more versatile. A minor gripe was that I couldn’t find a modifier key to lock the frequency while dragging up/down; you can of course use the lower section’s controls to achieve the same thing. And while you can use the Alt/Option‑click command to snap a node back to its zero position, I’ve sometimes wished for a similar way to snap the Tonal node to the Transient one, or vice‑versa (irrespective of whether that position is zero). But, as I say, these are small details, and on the whole it works very smoothly.

Split Siding Away

As well as the EQ, each band offers you a means of adjusting the stereo image. At the bottom‑left of the GUI you can switch the lower section to display these controls instead of the EQ ones. (They don’t take up a huge amount of space, and I wonder whether it could be useful to have the option to display both control sets simultaneously?) Essentially, you have the option of panning each path left or right, or of changing the Mid‑Sides balance of either path, or both of course. You can also pan/widen the two paths globally (all bands at once).

This might seem a curious facility to include on an EQ, but it’s potentially very useful here, simply because of the transient‑based split. For instance, by panning the Transient and Tonal elements of a mono acoustic guitar recording, I was able to create a lovely, natural‑sounding sense of stereo width — it’s nice to be able to do this using the ambience already baked into the recording rather than relying on artificial reverb — in a way that didn’t seem to cause problems for mono playback.

The lower section view can be switched to reveal panning and M‑S width controls for the two paths in each band.

Split: The Difference?

So far, I’ve written mostly about SplitEQ’s functionality, but not a huge amount about what you might actually want to use it for, and there’s a vast range of potential applications. While we often think of transients as being the most fleeting of moments, and maybe comprising mostly noise, SplitEQ allows you to define fairly precisely what constitutes the ‘transient’, and depending on the split settings it can be surprising just how much difference boosting or cutting at any given frequency in one path or the other can make.

Perhaps the best way I can convey this in words is to describe some examples, and I’ll start with vocals. By boosting only in the Tonal path of a lead vocal part you can change the overall brightness without bringing up much in the way of sibilance or lip smacking. Alternatively, you can apply that boost like a regular EQ, and then pull down any unwanted transient details that boost brings up without compromising that overall lift. I can’t think of any other EQ that can do that; I’ve used multiband transient shapers and dynamic EQs to achieve similar ends but the former is a blunter tool and the latter requires more fine‑tuning. Alternatively, if you have layer upon layer of background vocals, I find you can sometimes get an unhelpful build‑up of consonants and mouth noises that undermine the sense of timing. Pulling down distracting frequencies only in the Transient path can be a wonderfully subtle way to address that.

By boosting only in the Tonal path of a vocal part you can change the overall brightness without bringing up much in the way of sibilance or lip smacking.

Drums are an obvious candidate for any transient‑based process, of course, and SplitEQ can be really effective here. For example, my ears often seem to be offended by too much of the 200Hz region of an acoustic kit’s kick drum, and I almost always end up with a fairly narrow cut somewhere around that area. But it’s rarely a perfect solution. The ability to pull down only the sustain of the kick at that frequency — or to pull it down more than I do the transient — makes it easier to achieve what I want with less collateral damage. Higher up the spectrum, a boost to only the Transient path can be great if you need to bring out more of the beater click without raising unwanted content spilling onto the mic from other drums.

Drums, toms and snare can be manipulated in similar fashion. In particular, if you judge the split settings nicely, it’s an incredibly useful tool for pulling annoying resonances out of tom‑toms without robbing the initial hits of their impact. But SplitEQ also has plenty of potential for use on mixed drum signals too, whether they be drum loops or the drum bus in a mix. For example, I often use tape emulations on the drum bus but this can often have an undesirable softening effect on the cymbals. I really appreciate the control that SplitEQ gives me when EQ’ing ‘into’ the tape emulation, to manage that side‑effect.

Perhaps less obviously, SplitEQ can also be a really helpful tool when it comes to EQ’ing one source to create space for another, and not least in those regions where instruments often tend to clash with vocal parts. Take that 2.5‑ish kHz area in electric guitars, for example, which can help them cut through a mix (particularly when heard in mono, which often leaves impressive, aggressive wide‑panned guitars sounding rather duller than in stereo) but can sometimes also mask details in a vocal that are important to the lyrical intelligibility. If you get the separation/overlap between the Transient and Tonal paths right, you’ll find that you can often either boost the Tonal path of the guitar in that area, making it stand a bit prouder in the mix without suffocating important vocal consonants, or, if the guitar is present enough already, you can pull down just enough of the Transient that you can hear the lyrics more clearly without unduly compromising the guitar. While I’m on the subject of guitars, SplitEQ can be great for ‘de‑quacking’ piezo pickups or controlling unwanted ‘boom’ in acoustic guitar recordings.

Finally, while I don’t consider myself a ‘mastering engineer’ as such, I have tried a spot of DIY mastering with Split EQ and am convinced it could have applications in this role too, particularly for those projects which need a little more by way of corrective work, whether that be taming boomy bass instruments, or lifting up the high end without highlighting unwanted transient details or making things sound too ‘brittle’.

Verdict

I could discuss more examples but these ones should provide enough of a picture. The smoothness and precision of the transient detection process is impressive and, while the source presets are well judged, for the best results it’s important to use the path Solo buttons and fine‑tune the split by ear. Do that and, for corrective work, it’s often possible to get away with much less radical moves than when using a conventional EQ. Meanwhile, when it comes to gentler tonal shaping there’s oodles of potential, but I’ve particularly appreciated how smooth a high‑end boost it can deliver, and the per‑band, per‑path pan and width controls are icing on the cake. It won’t replace my preferred EQ (which supports Mid‑Sides and dynamic bands, among many other features) but it will join it, and I’m certain that I’ll be reaching for SplitEQ on pretty much every mix for the foreseeable future.