An Introduction to Limiters (and How to Use Them)
In this introduction to limiters, dive into the nitty gritty: when to use them, how to use them, and how to recognize when you’re over-processing.
This article is all about limiters—when to use them, how to use them, and how to train your ears to tell if they’re mangling your original sound source. But first, we need to define what a limiter actually is.
In this piece you’ll learn:
Want to try what you learn? Experiment with limiters in
Ozone 11 Advanced
iZotope Music Production Suite Pro: Monthly
What is a limiter?
A limiter catches the loudest peaks of an audio source and applies brick wall compression that prevents it from exceeding the clipping point of 0 dBFS. Limiters are used to increase percieved loudness by increasing the quietest parts of an audio signal while preventing the peaks from clipping. Limiting is the final and perhaps the most essential steps in mastering a song which allows you to turn up the mix to commercial loudness without distorting.
When to use limiters
You may find yourself using limiters on peaky instruments that otherwise sound fine (metal drums come to mind) when mastering. Vocals, which can also spike suddenly, are subjected to limiting here and there as well (Nectar Pro sports a switchable limiter).
Still, the majority of limiter-talk tends to center on the master bus, and this makes sense: Limiters are ubiquitous at the end of the mastering chain.
Think of the limiter as a bouncer, standing just outside the door of a club, keeping harsh digital overs outside of the proceedings, and doing so with the force of a brick wall, which brings us to…
We often call such processors “brickwall limiters,” as their implementation keeps the signal from breaking through a predetermined ceiling (maybe we should call them “brick-ceiling limiters”). Usually the desired ceiling is somewhere between -1 and -0.3 dBFS—the digital ceiling, beyond which you get distortion.
These limiters operate with infinitely high ratios, always implementing a delay in the signal in order to see any peaks coming down the pike. Sometimes you can tweak how far the limiter peers into the distance with a dedicated lookahead parameter. Sometimes this parameter is off-limits to the user.
Whichever kind of limiter you have, know that it may not protect against clipping, and your metering might not show the clipping unless it’s designed to display these ‘true peaks’.
Some digital limiters aren’t configured, out of the box, for catching the peaks that might occur between digital samples (intersample peaks). For that, you can switch to a mode specifically programmed to catch intersample peaks (commonly referred to as “ISP” or “True Peak”).
True peak operation
Within the last few years, true-peak technology has become de rigueur for any modern limiter to compete—and many limiters now offer it as a selectable option.
There is, of course, a more controversial point to make: yes, many mastering engineers use true-peak limiters—but plenty of engineers dislike ISP limiters, claiming their sound affects the material in a deleterious, adverse way.
Some would rather lower the output ceiling to ensure nothing got past 0 dBFS than submit a louder master with a true-peak limiter deployed. Others don’t even care about true-peak distortion at all.
Indeed, take your favorite tunes from the last five years and put them through iZotope RX. Click on Waveform Stats. I can guarantee that many of your favorite tunes easily clip past 0 dBTP.
Personally, I go on a case-by-case basis, depending on the needs of the client and the needs of the song. On a recent EP I mastered for the artist Peach Face, one tune didn’t take to ISP limiting so well, while another was improved by the addition of the ISP setting on Ozone.
Whether or not you want to use true-peak limiting is up to you or your mastering engineer. Strictly speaking, I can tell you this: use true peak whenever you want to make absolutely sure there is no clipping when your audio is run through a D/A converter.
How to set up a limiter
To set a limiter, add it as the last effect on your mastering signal chain and dial in the settings while listening to the loudest part of your track. Limiters typically have three main controls: threshold, release, and output ceiling.
Threshold: the threshold indicates the level at which the limiter applies brick wall compression. Set the limiter's threshold just under the audio signal's peaks until you get around -2 to -3 dB of gain reduction. If you’re using a stock limiter, any more than that can introduce noise and unwanted artifacts. The Maximizer module in Ozone, however, has a Magnify Soft Clip feature that allows you to boost loudness while maintaining the highest fidelity.
Release: the amount of time it takes for the limiter to stop after the signal falls below the threshold. Ozone let's you select between different release modes that automatically adjust the limiter's release times to your audio's transients. Its IRC (Intelligent Release Control) technology lets you boost the overall level of your mixes without sacrificing dynamics and clarity. It allows for the most aggressive limiting by using an advanced psychoacoustic model to intelligently determine the speed of limiting that can be applied to the incoming signal before producing distortion that is detectable to the human ear.
Output ceiling: establishes the maximum level the limiter will output. You want your track to sit around 0 dB without clipping so the first thing you want to do is set your output/ceiling level between -0.3 dB and -0.8 dB. This will set the brick wall threshold just below the clipping point.
Ozone 11 Advanced
To better hear how these work, try implementing the exercise we’ll outline in just a moment.
The rhythm of algorithms
These days, limiters often employ selectable algorithms to better suit your individual music, some of which deploy sophisticated multiband technology under the hood. Take some of Ozone’s venerated IRC algorithms, for example.
IRC stands for “Intelligent Release Control,” and to some extent, IRC will preserve the dynamics and overall clarity of your mix. Now, with so many different IRC settings—four overall, with separate “styles'' in IRC III and IRC IV—it can be tough to know which is best for the music at hand. Luckily, most companies offer manuals to explain their algorithms, and iZotope is no exception, providing a detailed explanation of IRC in their documentation.
Where does multiband come in? That would be IRC IV, which uses a multitude of bands, rather than three or four, as you would’ve seen in the conventional, multiband limiters of yesteryear. IRC IV’s bands are split up in a psychoacoustic way to better serve the material at hand, resulting in a more transparent and natural effect.
All of this is well and good, but how do you actually use a limiter to your advantage? After all, if set wrong, limiters can introduce their own distortion, as well as unpleasant side effects on the groove.
How to hear the difference in 4 easy steps
Here’s a step-by-step guide for tuning your ears to the sonic specificities a limiter imparts on your mix. I usually go through some iteration of this process every time I demo a newly-released limiter.
1. Link the limiter’s input and output as you tweak
Many limiters allow you to link the input gain or threshold control with the output ceiling, so that as you push one, the other comes down in level. This way, your ears won’t be fooled by a satisfying jump in loudness, and you can better judge the moment you’ve gone too far.
Ozone Pro has a linking feature that pairs the threshold slider to the ceiling. Turn that on as you pull the threshold down. At some point, you’ll hear audible distortion. Stop here and pull back until you don’t hear the distortion anymore.
Now you’re ready for step 2.
2. A/B between gain-matched bypass and limited signal
Compare the limited signal with its bypassed variant, carefully noting the differences in timbre. Does the track feel narrower with the limiter on? Are the transients pillowy, softer, or otherwise altered? Does the groove of the whole piece seem different? Note your generalized findings, and write them down if that helps you (it helps me with internalization).
3. Set up a delta test
This is where things get fun—and depending on the limiter, a bit tricky. We’re defining “delta” here as the real-time difference between your limited mix and bypassed version. It’s similar to soloing the esses of a de-esser (where you’d only hear the ess), or the noise output of a de-noiser (where you’d only hear what it’s removing).
Isolating and soloing these artifacts will call attention to them. Hearing them apart from the mix will tune your ears to what’s missing from your limited material once you switch it back on.
It’s a bit like training yourself to hear the effects of MP3 encoding using Ozone’s codec simulator: Once you hear what the encoding takes out of the original mix, those weird, lossy qualities of MP3s reveal themselves. The same general principle underscores this exercise.
Some limiters offer specific delta parameters, but not all do. Luckily, you can make your own with some clever routing, multiple instances of a limiter, and a plug-in that flips polarity.
Please note: different DAWs handle latency in different ways, so you may need to experiment for your own purpose in order to create a true A minus B listening test. Luckily, the workflow is roughly the same in Logic Pro X and Pro Tools, two of the more common DAWs.
First, kill the output of your source track, selecting “no output” (or whatever variant of that terminology your DAW provides) in the output section. Next, route the source to two auxiliary channels on two sends. Duplicate the limiter to both auxiliary channels, and bypass the limiter on your source track. On the second channel, flip the limiter itself into bypass.
Logic users take note: Do not use your DAW’s global bypass, but rather, the plug-in’s own bypass parameter.
Next, load any utility plug-in after the limiter that flips the phase of both the left and right channel.
Logic users, I cannot stress this enough: it’s essential for you to have both limiters instantiated for the trick to work, or else latency will be an issue, and you won’t hear what you’re supposed to hear—the sporadic limiting of your mix, usually on the kick and snare drums, when the limiter is audibly pushing down on signal.
A final note on delta tests: there are plug-ins that can actually set up delta tests for you, so you don’t have to go through all this routing yourself. Some of them are exceedingly cheap as well.
By way of example, here’s a snippet of music:
Limited Piece of Music
And here’s it’s delta/difference signal:
Limited Music's Delta Signal
4. Go back to the A/B comparison
That’s right: Go back to step two and compare the limited mix to its bypassed counterpart. Now that you’ve identified, heard, and internalized what the limiter is taking away, you should be able to hear the differences between the two more clearly. Start moving parameters around, and you’ll find you’re hearing what they’re doing; less guesswork will go into tweaking the limiter’s time-constants and stereo interdependence options, as your ear has tuned itself to what the limiter is taking away.
It’s Your Turn to Experiment with Limiters
Limiters are getting better and better all the time. Some are quite revolutionary—Ozone, for instance, employs an assistant which intelligently sets your limiter, and goes farther by placing a dynamic equalizer before the limiter which helps achieve an even more transparent operation.
The larger point is this: As limiters are changing all the time, it’s important to understand their basic functions and how you can use them in mixing and mastering to get the most out of your audio.