What Is The Difference Between Audio Compression & Limiting?
Compression and limiting are two critically important and somewhat similar processes in the trade of audio engineering. Understanding the difference between the two will help us choose the proper tools for various tasks in audio mixing.
What is the difference between audio compression and limiting? Dynamic range compression and limiting both work on the same principle of reducing an audio signal's dynamic range. Limiting is compression with a very high (often infinite) ratio. Compressors reduce signal levels above a set threshold by a ratio. Limiters are designed to set a maximum output level.
In this article, we'll discuss the differences and similarities between two of the most important signal processes in audio: compression and limiting. We'll consider a few examples of compressors and limiters to improve our comprehension and look at the typical applications of each.
What Is Dynamic Range Compression?
As was discussed earlier, compression and limiting are considered dynamic range compression effects/processes.
This means that both processes work by reducing the dynamic range of an audio signal. The dynamic range is defined as the difference or ratio between the loudest part of the signal and the quietest part of the signal and is generally measured in decibels (dB).
To reduce or “compress” the dynamic range, compressors and limiters both act by attenuating only the loudest parts of the signal.
The two key questions, then, are:
- What constituted the loudest parts?
- By how much should the loudest parts be attenuated?
The “loudest parts” are defined by the threshold of the compressor or limiter. The amount of attenuation is defined by the ratio of the compressor or limiter.
What is the threshold of a compressor? The threshold of a compressor is a set amplitude limit that dictates when the compressor will engage and disengage. As the input exceeds the threshold, the compressor kicks in (with its given attack time). As the input drops back down below the threshold, the compressor disengages (according to its release time).
What is the ratio of a compressor? The ratio of a compressor compares the number of decibels the input signal is above the threshold to the number of decibels the output signal is above the threshold. In other words, it is the relative amount of attenuation the compressor will apply to the signal.
There are other parameters to be aware of when discussing dynamic range compression and limiting, but the threshold and ratio are the two major factors.
In case you're wondering, the other noteworthy parameters are as follows (I've added links to informative articles on each parameter):
- Attack time: the time it takes the compressor/limiter to apply its full attenuation after the sidechain signal surpasses the threshold.
- Release time: the time it takes the compressor/limiter to disengage once the sidechain signal drops below the threshold fully.
- Knee: the transition curve of the compressor/limiter at and near the threshold.
- Makeup gain: the gain applied post-compression/limiting to bring the signal back up to its original peak level (if need be).
- Lookahead: the amount of time the program signal is delayed relative to the sidechain signal, effectively allowing the compressor to react before the program would otherwise trigger it.
Compression Vs. Limiting Basics
The major difference between dynamic range compression and limiting is the ratio.
Once again, the ratio refers to the ratio of the input level above the threshold to the output level above the threshold. In other words, as the ratio increases, so too does the attenuation above the threshold.
Limiting, as we've discussed, is essentially compression with a very high ratio. In modern and digital limiters (including plugins), the ratio is often ∞:1.
However, a compressor with a ratio above 10:1 can often be considered as a limiter or, at the very least, a compressor capable of “limiting”.
While a compressor will reduce the gain of the program audio signal above the threshold, a limiter will virtually prevent any signal level above the threshold.
Let's consider input/output graphs for a compressor with a 2:1 ratio (left) and a limiter with ∞:1 ratio (right).
For every 2 dB above the threshold, the compressor will only output 1 dB above the threshold.
With a ratio of ∞:1, the limiter will not output any level above the set “threshold”, which is often referred to simply as the output level of the limiter.
In order to clamp down and ensure that the output never surpasses the set maximum, a limiter must have a very fast attack time (ideally 0 ms) and/or an appropriate lookahead time. This type of limiter is known as a brickwall limiter.
The way in which a compressor and limiter will act upon an audio signal can be visualized in the image below. The threshold is drawn as the dotted red line:
Note that the compressor reduces the amplitude of the original signal above the threshold but still allows the signal to surpass the threshold. The limiter, conversely, limits the amplitude at the defined “threshold”.
In both cases, any signal below the threshold passes without any attenuation. Of course, attack time and release time would play a role in the resulting shape of the audio signals. However, for simplicity's sake, I've left them out of the simplifier image below.
This is the basic idea behind dynamic range compression. However, there is more to know about compressors and limiters. For example, the way the gain reduction element/circuit is fed is different between compressors and brickwall limiters.
To expand on our knowledge base, let's move beyond the basics and discuss compressors and limiters more thoroughly, with a keen focus on the factors that make them different.
What Is A Compressor?
We've already discussed how a compressor compresses an audio signal above a defined threshold by a defined ratio.
Though any ratio could be used, compressors often have a ratio below 10:1.
Some compressors have program-dependent ratios, meaning that the ratio is dependent on the sidechain signal level. In other words, the knee is rather soft, and the threshold is somewhat ill-defined.
Many variable-mu/tube compressors have program-dependent ratios.
The Retro Instruments Sta-Level Tube Compressor is an example of a program-dependent variable-mu/tube compressor without threshold or ratio controls.
Other compressors have adjustable ratio controls that go to and even above 10:1.
Above 10:1 and below brickwall limiting (∞:1) is a sort of grey area when compression could be considered as limiting.
Optical compressors, VCA compressors, diode bridge compressors, PWM compressors and FET compressors will often have adjustable ratios.
The Solid State Logic XLogic G Series Compressor is an example of a VCA compressor with an adjustable threshold of ±15 dB and 3 selectable ratios (2:1, 4:1 and 10:1).
The FabFilter Pro-C 2 is a digital compressor plugin with an adjustable threshold from -60 dB to 0 dB and an adjustable ratio from 1:1 to ∞:1.
A compressor may or may not also have an adjustable threshold control. Suppose a compressor doesn't have a threshold control (this is common with variable-mu/tube compressors and other program-dependent designs). In that case, it will likely have an input control to adjust the sidechain level that drives the gain reduction circuit.
Put differently, in many cases, the threshold and ratio of a compressor can be adjusted to compress the audio signal by varying degrees. In other cases, the threshold and/or ratio cannot be adjusted, so the audio level must be adjusted before the gain reduction circuit.
Note that compressors (and limiters) will have a program audio signal and a sidechain signal. The program audio is the audio that effectively gets compressed, and the sidechain signal is the signal that controls the compression.
In most cases, the sidechain signal is taken from the program signal directly before the gain reduction circuit (feed-forward design) or directly after the gain reduction circuit (feedback design). Alternatively, some compressors allow external sidechain signals to control the compression of the program/input audio signal.
Let's look at some block diagrams to improve our understanding.
Recap On Compressors
A compressor can be defined by the following statements:
• Generally has a ratio below 10:1, though higher ratios can also be considered to be compression.
• Has a defined threshold, whether it's adjustable or not.
• Uses a sidechain (internal or external) to control the gain reduction.
What Is A Limiter?
A limiter is a special type of compressor. As was previously discussed, a limiter may be defined as a compressor with a “high” ratio above 10:1. This was particularly true in the early days of compressors and limiters when digital audio and brickwall limiting were not yet invented (or at least commonplace).
This can be seen in the iconic Empirical Labs EL8 Distressor Compressor/Limiter, which has the following selectable ratios: 1:1, 2:1, 3:1, 4:1, 6:1, 10:1, 20:1 and “Nuke”.
It is also the case with the legendary Universal Audio Teletronix LA-2A Leveling Amplifier. This unit has a program-dependent input/output curve for both the compressor and limiter modes. The limiter will be more aggressive.
However, in modern-day audio technology and the age of digital audio, the term “limiter” generally refers to hard limiting or “brickwall” limiting. As we've discussed, this style of limiting is quite a bit different in application to “regular” compression.
Modern limiters effectively cap the maximum output level of a signal. Ideally (with an attack of 0 ms and a ratio of ∞:1), a limiter will ensure a signal never exceeds a set amplitude.
Sonically, limiters are significantly harsher than compressors and are most commonly applied as “safety devices” on mastering and final output stages in live sound, recording and broadcast situations. In many cases, these limiters are designed as software.
Limiters are designed primarily with a threshold and output level control. Like compressors, they will also likely have some combination of time controls (attack and release) and lookahead, along with optional DSP parameters.
In some designs, the threshold is adjustable. This can be seen in the popular Waves L3 Ultramaximizer.
Other limiters have a set threshold and gain stage that feeds into the threshold. This can be seen in the superb FabFilter Pro L2.
Virtually all limiters will have an adjustable output level control that will effectively act as the maximum peak output of the limiter.
Recap On Limiters
A limiter can be defined by the following statements:
• Generally has a ratio above 10:1, and brickwall limiters have a theoretical ratio of ∞:1.
• Has a defined threshold, whether it's adjustable or not.
• Uses a sidechain (internal or external) to control the gain reduction.
When To Use A Compressor & When To Use A Limiter
Now that we understand the differences between audio compressors and limiters, let's consider the applications of both. Though the two are based on the same general working principle, they are distinguishable by their applications in audio mixing.
Compressor Applications
Compressors are used much more often than limiters in audio mixing.
Applications for compressors include, but are not restricted to, the following:
- Maintaining a more consistent level across an audio track.
- Improving sustain of audio signals.
- Uncovering nuance in an audio signal by bringing the quiet parts up relative to the loud parts.
- Gluing a mix together (mix/master bus compression).
- Shaping transients.
- De-essing (via multiband compression).
- Sidechain “pumping” and/or rhythmic effects.
- Adding subtle saturation.
Limiter Applications
Limiters are much more specialized in audio applications.
High ratio (non-brickwall) limiters can be used to varying success in the compressor applications mentioned above but are perhaps more commonly applied to the following:
- Parallel compression (on the over-compressed track/bus).
- Talkback microphones and other communication channels.
- Moving elements to the back of the mix.
- Tone shaping.
Applications for hard limiters or “brickwall” limiters include, but are not limited to, the following:
- Keeping a final output below the 0 dBFS ceiling of digital audio (keeping the signal from becoming digitally clipped at the output). This is useful in every digital audio mix.
- Clipping the audio (ideally as an effect).
- Controlling sharp transients of individual tracks.
Here's one of my videos explaining how to use limiting to get “competitive” levels in a mix for monitoring and referencing before mastering:
Call To Action!
Select a track that could benefit from dynamic range control. Experiment with a compressor, adjusting the ratio to reach limiting (10:1 ratio at a minimum) and A/B how it sounds versus a lower ratio (say, 2:1 or 4:1). Note how the peak level may or may not be different depending on the amount of gain reduction.
Consider, for yourself, what production and mixing situations may benefit most from regular compression versus those that would benefit most from limiting.
Have any thoughts, questions or concerns? I invite you to add them to the comment section at the bottom of the page! I'd love to hear your insights and inquiries and will do my best to add to the conversation. Thanks!
Related Questions
Should you put a limiter on every track? Limiters are not required or even preferable on every track in a mix. In digital recordings, limiters are generally reserved for mix/master bus processing to keep levels below a defined maximum value (typically 0 dBFS). Limiters (both digital and analog) can also be used sparingly on tracks with significant transient to help smooth out levels.
How much headroom do you leave for mastering? Although mixing and mastering continue to become intertwined, if you do happen to send off a mix to a mastering engineer, a good rule of thumb is to leave 3 to 6 dB of headroom before mastering. Be sure to communicate what the deliverables ought to be with the mastering engineer you'll be working with.
