You are here

Using Transient Processors

The granddaddy of transient processors, the SPL Transient Designer hardware outboard unit, in its current MkII incarnation.The granddaddy of transient processors, the SPL Transient Designer hardware outboard unit, in its current MkII incarnation.

Attack‑enhancing plug‑ins can transform your mix — but they need to be used wisely!

One of the more common reasons for processing a track at mixdown is because you want to reshape its attack in some way. Your kick drum might feel like it lacks punch, say, or you want more ‘smack’ from your snare. Alternatively, maybe your percussion feels too spiky and upfront, or your piano part feels overly hard‑edged and aggressive.

Now, a lot of project‑studio owners instinctively reach for EQ or compression to address these kinds of issues, but those tools aren’t always well suited to the task. So you might try to boost your kick drum’s low midrange to add punch, but discover that this makes the drum’s tone horribly woolly into the bargain. Or you might cut some high end from the percussion to soften its attack, only to find that it now sounds dull and congested. Many producers have also noticed that slower‑attack compression can enhance a snare drum’s onset level spike (often referred to as its ‘transient’) by allowing it to zip through the processor before the gain reduction has a chance to react. However, because compressors work according to a fixed threshold level, this trick quickly becomes rather inconsistent with drum parts that have any dynamic variation to them — typically, louder hits will get more attack enhancement than quieter hits, which may not be what you want! Likewise, fast‑attack compression is often used to reduce attack, but if you try this kind of treatment on a hard‑edged piano part, you’ll rarely be able to bring the compression threshold down far enough to soften the transients of lower‑level notes without unconscionably level‑pumping or distorting the rest of the instrument’s sound.

This kind of processing incurs far less ‘collateral damage’ on the rest of the signal, because you can cut/boost just the transients, without affecting the sound’s overall timbre or dynamics.

Fortunately, there are specialist processors that can avoid many of these problems by approaching transients in a different way: not in terms of frequency response or signal level, but more appropriately in terms of how fast they cause a signal’s level to change. This kind of processing incurs far less ‘collateral damage’ on the rest of the signal, because you can cut/boost just the transients, without affecting the sound’s overall timbre or dynamics. And because a signal’s rate of level change is independent of its absolute level, transients are detected and processed the same irrespective of whether they’re loud or quiet, which means you get more consistent attack shaping when working with dynamic musical performances.

The granddaddy of transient processing was undoubtedly SPL’s original 1998 Transient Designer rack unit (shown above), and the company continue to manufacture the product to this day (albeit in an updated MkII form), as well as offering a plug‑in version. In the software domain, however, transient‑processing functionality is now actually built into many DAW systems, and even if it’s not available in your recording software, there are plenty of high‑quality cross‑platform freeware plug‑ins available, such as Kilohearts’ Transient Shaper, Flux’s BitterSweet, or Auburn Sounds Couture.

If your DAW software doesn’t include transient processing by default, or you’d just like to try out some different algorithms, there’s plenty of excellent cross‑platform freeware available, such as Kilohearts’ Transient Shaper, Auburn Sounds’ Couture and Flux’s BitterSweet.If your DAW software doesn’t include transient processing by default, or you’d just like to try out some different algorithms, there’s plenty of excellent cross‑platform freeware available, such as Kilohearts’ Transient Shaper, Auburn Sounds’ Couture and Flux’s BitterSweet.

Practical Applications

But enough of the theory. How can you get the best results out of these dedicated transient processors in practice? Well, the good news is that the main control, often labelled something like Attack, is extremely simple. Turn it up to get more attack. Turn it down to get less attack. And if all you want to do is add a bit of snap to your drum loop, or remove unpleasant spikiness from close‑miked hand claps, then you may need do nothing more than just twirl that control and trust your ears. Indeed, for at least half the transient processors I use while mixing, that’s really all there is to it!

That said, a big part of the battle is just realising that transient processing might be able to help, because its usefulness extends well beyond obvious drum‑enhancement roles. I regularly turn to it for smoothing out things like cymbal stick noise on drum overhead mics, pick noise on acoustic guitars, string slap on upright bass, or clicks and crackle on samples taken from vinyl. I’ve even used it on lead‑vocal and voiceover tracks to combat abrasive ‘t’ and ‘k’ consonants, or those little snicking noises you get when the singer’s mouth’s got a bit dry.

For programmed electronic music, transient reduction can be really useful for smoothing out complex background‑pad presets that include percussive step‑sequenced or arpeggiated elements that would otherwise distract the listener’s attention from more important foreground parts such as a topliner’s verse. If you like combining multiple sampled breaks, or underpinning breaks with programmed beats, then it’s not unusual for some of your layers to be more important for their timbre and sustain rather than for their attack. In such cases, you can sometimes use transient reduction on those textural layers to reduce any rhythmic flamming between them and the other layers, thereby giving a cleaner and punchier attack to the overall drums submix.

Headroom Matters

Pushing The EnvelopeSome DAWs already have specialist transient‑processing plug‑ins built in. For example, Steinberg’s Cubase has its Envelope Shaper and Cockos Reaper has JS Transient Controller.Some DAWs already have specialist transient‑processing plug‑ins built in. For example, Steinberg’s Cubase has its Envelope Shaper and Cockos Reaper has JS Transient Controller.From a technical standpoint, though, you should be aware that boosting a sound’s attack will often also increase its peak level, especially where you’re heavily processing drum tracks. In modern DAW software this needn’t be a problem, just so long as you make sure you’re not inadvertently clipping your master bus or any of your system’s analogue outputs (and thereby unintentionally introducing distortion). However, if you’re using any downstream dynamics processors (a compressor on the drums bus, say), a big increase in peak levels can significantly change how they respond.

By way of acknowledgement that headroom can become a concern, some transient processors now include a switchable limiting or clipping section to contain unwanted peak‑level overshoots. Now, it might seem a bit daft to first boost the level of a drum’s transient, and then immediately limit/clip that level peak. But it’s important to realise that limiting and clipping won’t completely negate the effect of a transient boost.

For a start, there may be some quieter hits that remain within your available headroom, so won’t trigger the clipping/limiting process. But even those hits that do come up againt the safety endstop won’t lose all the benefits of the transient boost. As I see it, this is because a transient’s peak level isn’t the only thing the processor changes — it also increases the rate at which the level rises as the transient develops (what I like to call its ‘rise time’). So even if a limiter or clipper curtails a boosted transient’s peak, the rise time will still be faster than it was, and the drum hit will thus appear to have more attack nonetheless. Mind you, a limiter will typically counteract the subjective effect of a transient boost much more strongly than a clipper will, so you may prefer to use the clipping option if you’re after maximum subjective attack and are less concerned with the unavoidable tonal side‑effects of the resultant clipping distortion.

Refining The Transient

Another facility you’ll see on many transient processors is some way to adjust the shape of the processing envelope. At its simplest, this will just be a single Speed control or time slider, which determines how quickly your desired gain boost/cut kicks in once a transient is detected, and how quickly the gain resets once that transient’s passed by. More fully featured plug‑ins, however, may have all sorts of rather inscrutable envelope options to choose from. Don’t let that intimidate you, though, because whatever settings are on offer you’ll usually find that they just give a broad continuum between faster envelope actions (which tend to add a sharper attack ‘spike’) and slower envelope actions (which typically provide a better sense of low‑end weight). Once you recognise that fundamental trade‑off, it’s pretty straightforward to surf the available settings to find the ones that give you the best balance between ‘snap’ and ‘punch’. If I’m wanting to soften transients rather than boost them, though, then I’ll almost always choose the fastest gain‑change envelope, simply because I find that it’s the sharp onset and upper‑spectrum content of overbearing transients that usually make them most unpleasant to listen to, and a faster gain‑change envelope better targets those aspects of the sound.

On occasion, however, I do want more control over the tone of my transient processing. For example, I regularly encounter project‑studio multitracks where the programmer or tracking engineer has tried (unsuccessfully!) to salvage a gutless kick‑drum sound by overemphasising some upper‑spectrum attack component — for example the beater ‘click’ of a close‑miked acoustic kick drum. The real remedy for this is to boost the low spectrum of the drum’s onset transient, which is something that traditional transient processing can’t really do, no matter how much you play with its envelope settings. Happily, there are a couple of good workarounds here.

The first is to set up your transient processor as a send effect rather than just inserting it on the kick‑drum channel — in other words, to use the plug‑in as a parallel processor. In this configuration, you can boost the drum’s transient in the effect‑return channel, and then EQ that transient‑enhanced signal independently before mixing it back in with the unprocessed drum sound. So in this specific kick‑drum scenario, I might low‑pass filter the transient‑processed addition to beef up just the weightiness of the drum’s impact.

The second workaround is to get hold of one of the more powerful transient processors that allow you to tailor their processing to different frequency ranges — things like Oeksound Spiff, Boz Digital Transgressor, Melda MTransientMB, or iZotope Neutron Transient Shaper. In our kick‑drum example, those tools would let you boost just the low end of detected transients in order to implement a similar outcome.

For more precise frequency‑selective transient processing, you may want to investigate feature‑heavy spectral processors such as Oeksound Spiff, Melda MTransientMB, Boz Digital Transgressor, or iZotope Neutron Transient Shaper — or else consider using simpler transient processors in a parallel‑processing configuration.For more precise frequency‑selective transient processing, you may want to investigate feature‑heavy spectral processors such as Oeksound Spiff, Melda MTransientMB, Boz Digital Transgressor, or iZotope Neutron Transient Shaper — or else consider using simpler transient processors in a parallel‑processing configuration.

The down side of the parallel processing setup is that it’s only really useful for attack boosts, not attack cuts, and you also have to be careful that phase shifts incurred by your EQ don’t cause problematic comb‑filtering between the processed and unprocessed signals. (If they do, try a linear‑phase EQ instead.) But on the up side, with the parallel setup you can use more easily understood plug‑ins — which may also be freeware into the bargain!

Whichever tactic you go for, though, my kick‑drum example is only one of the potential applications for frequency‑selective transient processing. Indeed, many of the use cases I’ve already mentioned can be more effectively handled in this way. With acoustic‑guitar, for instance, you can usually rein in pick noise much more firmly (without unacceptable processing side‑effects) by focusing your transient‑reduction processing above 5kHz.

Transient Joy

Once you get a handle on what this nifty processing option has to offer, you’ll likely find yourself getting mix results quicker, without having to push your EQ and compression settings to extremes — so it’s definitely worth spending some time experimenting with the tricks I’ve mentioned here. One word of caution, though: transient boosts in particular can be strangely addictive, because they’ll make most instruments sound more upfront and engaging. But if you make everything upfront and engaging, then your mix will likely sound cluttered and lacking in any sense of depth. As with many things in life, moderation is a virtue.  

Threshold‑dependent Processing

In this article I’m discussing threshold‑independent transient processors, because these are the most common ones you’ll find these days. There are some transient processors, however, that actually do work according to a fixed level‑threshold — so whenever a drum signal exceeds the level threshold, a tiny gain spike is triggered to boost/cut what is usually the onset of the drum hit. Although this approach can create very similar attack‑adjustment results for unvarying drum samples, any drum part with musical dynamics can produce a rather hit‑and‑miss outcome here, so I’m not a huge fan of threshold‑dependent transient processing for general mixing purposes. I certainly wouldn’t spend money on such a plug‑in, as it’s very easy to recreate very similar functionality by setting up a fast‑acting gate as a send effect: whenever a drum hit exceeds the gating threshold, it lets though a short burst of drum transient which you can simply mix in to bolster the attack of the unprocessed signal to taste.

Sustain Enhancement

Alongside the Attack control of some transient‑enhancement plug‑ins you may also see a second knob labelled Sustain, and the implementation of the two is actually quite similar: whereas the Attack control operates on signal levels that are rapidly increasing (ie. transients), the Sustain control operates on signal levels that are rapidly decreasing (ie. decay tails). If you’re working with drum samples or drum close mics, such Sustain controls can be remarkably effective, dramatically boosting drum sustain, or (in cut mode) drying up unwanted resonances and room reverb. However, any gain change that’s applied will usually then reset after some preset period of time, and while this isn’t usually a problem with isolated drum hits, it can begin to sound rather unnatural with more complicated mixed signals where multiple musical components are sustaining across each other. So, personally, if I’m looking for serious sustain enhancement, I’ll usually turn to parallel compression instead, as this always seems to sound more musically appropriate to me.

Audio Examples & Online Resources

As usual, I’ve provided a selection of audio examples to accompany this article on the Sound On Sound website at https://sosm.ag/transient-processing. In addition, I’ve created a special resources page at https://cambridge-mt.com/misc/sos-transient-processing/ where I’ve posted a couple of hands‑on video tutorials illustrating the two parallel‑processing techniques mentioned in this article. Alongside those, I’ve also collected together links to a wide range of freeware and payware transient‑enhancement plug‑ins, in case you’re looking to supplement your DAW’s existing processors.