Feedback Vs. Feed-Forward Dynamic Range Compressors In Audio

The distinction between feedback and feed-forward compressors is one of those more advanced topics that we should ask about once we understand the essentials of audio dynamic range compression.

What is feedback compression? Feedback compression feeds the audio signal into the sidechain just after the gain reduction element. This compressor type reacts to the signal amplitude without anticipating.

What is feed-forward compression? Feed-forward compression feeds the audio signal into the sidechain before the gain reduction element. This compressor type anticipates the signal amplitude and adjusts the sidechain signal in advance.

In this article, we'll discuss the theory behind feedback and feed-forward compressors in greater detail and consider examples of each of these compressor types. By the end of this post, you'll have a more complete understanding of compression and its role in audio production/processing.

---

Table Of Contents

• What Is Compression?
• Feedback Vs. Feed-Forward Compressors
  • The Feedback Compressor
  • The Feed-Forward Compressor
  • Comparing Applications For Feedback & Feed-Forward Compressors
• Examples Of Feedback Compressors
  • Universal Audio 1176LN
  • Rupert Neve Designs 5254
• Examples Of Feed-Forward Compressors
  • Elysia MPressor
  • dbx 160A
• Examples Of Compressors With Feedback & Feed-Forward Topology
  • API 2500+
  • Great River PWM 501
  • Undertone Audio Unfairchild 670M II
  • Mäag Audio Magnum-K
• Related Questions

---

What Is Compression?

To fully understand feedback and feed-forward compression, let's begin by discussing compression in more general terms.

Click here to skip ahead to the section Feedback Vs. Feed-Forward Compressors.

Compression (more technically “dynamic range compression”) is a process in audio that reduces or “compresses” the dynamic range of an audio signal. Though the effect/process has plenty of uses in the world of audio, its main purpose is to make the difference between the highest and lowest amplitudes of the audio signal smaller.

In general, this means attenuating only the loudest parts of the signal.

In order to attenuate the “loudest parts” of an audio signal, two key questions must be answered:

• What are the “loudest parts”?
• By how much should the “loudest parts” be attenuated?

A compressor answers these questions with the threshold and ratio parameters, respectively.

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.

Other compressor parameters worth mentioning are the following (I’ve added links to in-depth articles on each parameter):

• Attack Time: the amount of time it takes for a compressor to engage/react once the input signal amplitude surpasses the threshold.
• Release Time: the amount of time it takes for the compressor to disengage (to stop attenuating the signal) once the input signal drops below the threshold.
• Knee: the transition point around the threshold of the compressor where the output becomes attenuated versus the input.
• Makeup Gain: the gain applied to the signal after the compression takes place (typically used to bring the peaks of the compressed signal up to the same level as the peaks pre-compression).

To attenuate the audio signal, a compressor must have a gain reduction circuit. The types of gain reduction circuits vary from compressor to compressor and largely define the “type” of the compressor in question. The most common compressor types include (I’ve added links to in-depth articles on each type):

• Diode bridge compressor: uses a diode bridge as the gain reduction element.
• FET compressor: uses a field-effect transistor as the gain reduction element.
• Optical compressor: uses an optical assembly (a voltage-controlled light source and light-dependent resistor) as the gain reduction element.
• PWM compressor: uses a pulse width modulator as the gain reduction element.
• Variable-mu/tube compressor: uses a remote cut-off vacuum tube as the gain reduction element.
• VCA compressor: uses a voltage-controlled amplifier as the gain reduction element.

For any of these gain reduction elements/circuits to attenuate the audio signal, they must be effectively “told” how to do so. The control signal for these elements is referred to as the compressor's sidechain signal.

The sidechain signal will typically go through some sort of level detector/rectifier (peak level, RMS level or otherwise) and come out as a variable DC control signal. The sidechain signal path will also manipulate the signal in order to achieve some or all of the aforementioned parameters (threshold, ratio, time controls, etc.).

Once effectively manipulated, the sidechain will control the gain reduction circuit and “tell” it how to compress the main audio/program signal.

Though external audio signals may be used as the sidechain, it's more commonly the case that the program audio is sent to the sidechain signal path in addition to the audio/compression signal path.

When the program audio is taken as the sidechain, the compressor's “split” point can be set up before or after the gain reduction circuit. This is where the design concepts of feedforward and feedback come into play, respectively.

Related article: The Complete Guide To Audio Compression & Compressors

---

Feedback Vs. Feed-Forward Compressors

Now that we understand the very basics of compression, let's get to the main part of this article.

As was mentioned previously, feedback and feed-forward compressor topologies refer to where, in the circuit, the sidechain signal comes from.

A feedback compressor feeds the audio signal into the sidechain just after the gain reduction element. This compressor type reacts to the signal amplitude without anticipating.

Here is a basic block diagram of a feedback compressor (note that the “sidechain” block houses the level detector/rectifier and any controls over the compressor parameters).

A feed-forward compressor feeds the audio signal into the sidechain just before the gain reduction element. This compressor type anticipates the signal amplitude and attempts to adjust the sidechain signal in advance.

Here is a basic block diagram of a feed-forward compressor (note that the “sidechain” block houses the level detector/rectifier and any controls over the compressor parameters).

Consider, for a moment, an ideal compressor: one that does not affect the audio signal whatsoever except by a defined amount of gain reduction is and when the sidechain signal surpasses the threshold.

In this ideal compressor, a feed-forward and a feedback design would act the same below the threshold. The input signal, output signal and sidechain signal would all be at the same level, and no gain reduction would take place.

Of course, we don't live in an ideal world, and the circuit itself will cause inherent colouration to the signal. Even with no gain reduction, the sidechain taken before the gain reduction circuit (feed-forward) would be different, even if only slightly, from the sidechain taken after the gain reduction circuit.

Now comes the more important question. What happens when the sidechain signal level surpasses the threshold and causes the gain reduction circuit to attenuate the signal?

The Feedback Compressor

It would seem that in the case of the feedback design, it would be the already compressed signal that would cause the compression in the first place. This is correct. As the program audio signal surpasses the threshold, it passes through the gain reduction circuit and the sidechain, in that order. The signal is also, of course, passed through to the output.

The sidechain then tells the gain reduction circuit to attenuate the signal, which is then fed back into the sidechain (and output). So the sidechain is constantly adjusting the amount of compression/attenuation it, itself, is experiencing.

You may be thinking that this would cause a significant delay in how a compressor will react. That's partially true. However, electricity moves nearly instantaneously and other than reactive components (capacitors, inductors, etc.), the feedback design can potentially react very quickly in its compression.

There will be some amount of lag relative to feed-forward designs. However, feedback designs, especially those in VCA, PWM, FET and diode bridge compressors, are fast-acting.

It stands to reason, then, that the type of compressor gain reduction circuit has a larger role in determining the overall reaction time of the compressor. Remote cut-off tubes and optical assemblies are relatively slow at reacting to their sidechain control signals. Therefore, variable-mu and optical compressors are inherently “slow” compared to most other types.

Of course, the most important factors when it comes to compressor reaction speed are the attack time and release time controls of the compressor in question (if the compressor has these adjustable controls.

The Feed-Forward Compressor

The feed-forward design, by contrast, feeds the sidechain with the same signal it feeds the compression circuit. In this case, the sidechain reads/detects the audio program signal pre-compression and creates a control signal (with the compressor's parameters) that will ultimately compress the same audio program signal.

With a feed-forward design, we can manipulate the detected sidechain control signal with the typical parameters (threshold, ratio, attack, release, etc.). In fact, because the sidechain signal is unaffected, it's to the compressor's benefit to have as many controllable parameters as possible.

Because the feed-forward sidechain can never adapt to the signal after its compression, it must rely on the control signal detector and the gain reduction circuit having the same input-to-output characteristics. This is hardly a simple proposition and generally requires a linear gain reduction circuit to go along with a linear detection circuit.

In other words, the control signal level must be proportional to the gain reduction level (ideally linear) so that the feed-forward sidechain will tell the compressor to attenuate the program signal properly.

This is relatively easy to achieve with VCAs (voltage-controlled amplifiers) and PWMs (pulse width modulators), which are, for the most part, linear. In fact, the popularity of the feed-forward design came about due, in large part, to the invention of the Blackmer gain cell VCA and the VCA compressor.

Diode bridge compressors could be designed with a feed-forward design, though they're rather complex to design.

Optical, FET and tube compression circuits are rather non-linear, so feed-forward designs typically do not serve them well. That being said, compressor designers are an ambitious bunch, and there are feed-forward designs for many non-linear gain reduction circuits (we'll discuss a few examples later).

Though both feed-forward and feedback compressors are capable of fast attack and release times (depending on the gain reduction circuit type), the feed-forward topology can act more instantaneously as it doesn't have to adjust to any signal that's already been compressed.

Therefore, feed-forward circuits are favoured for catching fast transients and for hard limiting. Hard limiting is the process where, ideally, absolutely no signal can be passed above a hard threshold (think of it as a compressor with an attack time of 0 and a ratio of ∞:1).

Feed-forward designs can approximate the idealization, whereas feedback designs, though fast, could never reach this instantaneous reaction time.

As an aside, it's important to note that early limiters (and even many limiters nowadays) were not at all the “brickwall” limiters we're all too familiar with today. The limiters often had high ratios (20:1) but did allow transients to pass through as they reacted to signals passing above the threshold.

As another aside, noise gates absolutely have to be feed-forward. If they were feedback, then they'd never open up once they closed. Noise gates effectively mute a signal once it drops below a defined threshold. A noise gate is to an expander what a hard limiter is to a compressor.

Comparing Applications For Feedback & Feed-Forward Compressors

Now that we understand what feedback and feed-forward compressors are, let's consider their uses.

In general, feed-forward compressors are preferred for hard limiting applications and their ability to compress fast and hard.

If a ton of compression is needed (perhaps for parallel/Manhattan-style compression), go for a feed-forward. Likewise, if transients really need to be tamed in a certain track, a feed-forward compressor can be an invaluable tool.

On the other hand, Feedback compressors are often chosen for bus compression and for applications that don't require a significant amount of gain reduction or a particularly aggressive style of compression.

That being said, other than hard limiting, either topology can be used effectively in practically any application so long as the compressor is versatile in its performance.

Many of the differences between feedback and feed-forward compressors can be summed up in the table below:

Feedback Compressor	Feed-Forward Compressor
Sidechain detector after the gain reduction circuit.	Sidechain detector before the gain reduction circuit.
Constantly re-evaluates and adjusts the amount of compression.	Relies on proportionality between the I/O characteristic of the level detector and gain reduction circuits.
Less potential for over-compression.	Greater potential for over-compression.
Less precise control parameters.	More precise control parameters.
Typically have fewer parameter controls.	Typically have all parameter controls.
Relatively more interdependence between parameter controls.	Relatively more independence between parameter controls.
Typical of: VCA compressors PWM compressors FET compressors Optical compressors Variable-mu compressors Diode bridge compressors	Typical of: VCA compressors PWM compressors
Cannot be used to achieve hard limiting.	Can be used to achieve hard limiting.

---

Examples Of Feedback Compressors

When learning a new topic in audio (and practically anything else), it's helpful to consider examples. Let us examine two specific feedback compressors to help deepen our understanding.

In this section, we'll discuss the following feedback compressors:

• Universal Audio 1176LN
• Rupert Neve Designs 5254

Universal Audio 1176LN

The Universal Audio 1176LN is a hand-built faithful reproduction of the legendary Universal Audio 1176 Limiting Amplifier, which was first released in 1967.

This FET compressor, like practically all FET compressors, is designed with feedback topology.

This is because FET-based gain reduction circuits only perform within a small transfer region in which the attenuation/amplification is non-linear. By constantly examining and adjusting the control signal, the FET can properly function as the centrepiece of the compressor in reaction to the sidechain control signal.

Rupert Neve Designs 5254

The Rupert Neve Designs 5254 is a stereo diode bridge compressor in a rack-mount form factor.

Like most diode bridge compressor designs, the 5254 has feedback topology.

The region in which compression can be achieved by varying the resistance of the diodes is very small and rather non-linear. Therefore, these relatively rare compressors will generally be designed with feedback topology. This allows for the consistent adaptation of the control signal to maintain proper compression.

---

Examples Of Feed-Forward Compressors

Now let's turn our attention to examples of feed-forward compressors. In this section, we'll be discussing the following compressors:

• Elysia MPressor
• dbx 160A

Elysia MPressor

The Elysia MPRESSOR is a stereo VCA compressor designed in the standard 19″ rack-mount form factor.

The MPressor, like many modern VCA compressors, utilizes a feed-forward approach to its compression. This allows the MPressor to react with superb precision and offers the unit tremendous flexibility. Controls include threshold, attack, release time, ratio, gain reduction limit, makeup gain and even EQ.

dbx 160A

The dbx 160A is another VCA compressor in a rack-mount design.

The dbx 160A is a classic compressor and has been making a name for itself since 1976. This feed-forward compressor was the first commercially available VCA compressor brought to you by none other than David E. Blackmer, the founder of dbx and inventor of the Blackmer gain cell (voltage-controlled amplifier).

The dbx 160A has rather minimal controls relative to modern VCAs, but it's a classic for a reason. It sounds incredible!

---

Examples Of Compressors With Feedback & Feed-Forward Topology

Before we wrap things up, let us look at a couple of compressors that have both feedback and feed-forward options. In this section, we'll be discussing the following compressors:

• API 2500+
• Great River PWM 501
• Undertone Audio Unfairchild 670M II
• Mäag Audio Magnum-K

API 2500+

The API 2500+ is a stereo VCA bus compressor built into a rack-mount form factor.

This compressor is impressively powerful and versatile. It offers the typical variable threshold, attack, ratio, release and makeup gain control. It also has 6 different tonal options, adjustable knee, API's patented thrust control (introduces a low-cut and high-boost equalizer into the side-chain detector circuit), and an option to choose feedback or feed-forward compression.

Stereo linking and parallel compression are also made possible in this highly functional unit.

Great River PWM 501

The Great River PWM 501 is a pulse width modulation compressor with blendable feed-forward and feedback compression, adjustable via the FF/FB mix control knob.

This particular compressor is unique in the fact that it can blend both feedback and feed-forward circuits together to achieve unique flavours of compression. This is made “easy” by the fact that the pulse width modulator attenuation is a well-defined function of its duty cycle. Therefore, the duty cycle percentage can be simultaneously influenced by the feed-forward and feedback sidechain control signals.

Undertone Audio Unfairchild 670M II

The Undertone Audio Unfairchild 670M II is a modernized recreation of the classic Fairchild 670 stereo variable-mu/tube compressor from 1959. It's an “exception to the rule” that states that all tube compressors are feedback.

This impressive feat of engineering effectively allows the non-linear (by nature) tube gain reduction circuit to be controlled by a feed-forward sidechain. It's truly a masterpiece, bridging the legendary sonic character of the Fairchild 670 with state-of-the-art modern circuit design.

There's too much to go through with this beast of a compressor, and I'd invite you to check it out for yourself and Undertone Audio's official website.

Mäag Audio Magnum-K

The Mäag Audio Magnum-K is a compressor/EQ with dual optical compressor stages. It's an “exception to the rule” that states that all optical compressors are feedback.

The main Magnum optical compressor offers both feedback and feed-forward operation, selectable via a switch on the front panel. The second K-comp optical compressor is set up in a typical feedback setup but is focused on the 3kHz region.

Like the aforementioned Undertone Audio compression, this Mäag Audio really pushes the envelope in terms of compressor technology, designing an optional feedforward circuit that effectively interacts with its optical assembly gain reduction circuit in a linear/proportional fashion.

---

Related Questions

What are the different types of audio compressors? The term “type” can have a few meanings so let's look at a few different “types of compressors.

In terms of circuit topology, compressors will generally fall into one of the following types:

• Variable-Mu (Tube) Compressor
• FET Compressor
• Optical Compressor
• VCA Compressor
• Diode Bridge Compressor
• Pulse Width Modulation Compressor
• Digital Compressor
• Compressor Plugin

In terms of how a compressor will perform when compressing an audio signal (and the typical tasks it will be set to do), we can think of the following types of compression:

• Multiband Compression
• Peak-Metering Compression
• RMS-Metering Compression
• Feedback Compression
• Feed-forward Compression
• Upward Compression
• Limiting Compression
• Parallel Compression
• Bus Compression

What does a limiter do in audio? A limiter is an audio processor that caps the maximum amplitude of a signal. The limiter will effectively limit the output amplitude to a set level and attenuate all input level peaks that exceed the threshold in order to maintain this output level. A limiter is essentially a compressor with a ∞:1 ratio.

---

Building out your 500 Series system can be a challenging task. For this reason, I've created My New Microphone's Comprehensive 500 Series Buyer's Guide. Check it out for help in determining your next 500 Series purchases.