My Take on Music Recording with Doug Fearn


March 01, 2024 Doug Fearn Season 1 Episode 89
My Take on Music Recording with Doug Fearn
More Info
My Take on Music Recording with Doug Fearn
Mar 01, 2024 Season 1 Episode 89
Doug Fearn

Send us a Text Message.

Compression and limiting are tools we use to modify the dynamic range of the music we record. In this episode, I present a brief history of where this technique came from, how it evolved over the years since the 1930s. I discuss the various ways that compression circuits work, from the variable-mu vacuum tube, to the optical compressor, the FET, the VCA, the PWM, and the digital limiters. I explain how these different approaches affect the sound in different ways.

There are often a lot of adjustments on a compressor, and I go through the most common controls you are likely to encounter and what they do to the sound.

Compressor and limiter effects on the music are subjective, so I try to give general guidance for getting the sound you want from your hardware and software devices.

Your ideas for future episodes are always welcomed. And your comments are useful to me. You can reach me at


Show Notes Transcript

Send us a Text Message.

Compression and limiting are tools we use to modify the dynamic range of the music we record. In this episode, I present a brief history of where this technique came from, how it evolved over the years since the 1930s. I discuss the various ways that compression circuits work, from the variable-mu vacuum tube, to the optical compressor, the FET, the VCA, the PWM, and the digital limiters. I explain how these different approaches affect the sound in different ways.

There are often a lot of adjustments on a compressor, and I go through the most common controls you are likely to encounter and what they do to the sound.

Compressor and limiter effects on the music are subjective, so I try to give general guidance for getting the sound you want from your hardware and software devices.

Your ideas for future episodes are always welcomed. And your comments are useful to me. You can reach me at


89           Compression                                                                      February 27, 2024


 I’m Doug Fearn and this is My Take on Music Recording

 Compression and limiting have become nearly universal tools in most music production. But where did this effect originate? And how does the equipment do it? Why are there so many different compressors? And when do you use compression and when is it better to avoid it?

 Back in the late 1920s, radio broadcasting was becoming a major force in society. It’s difficult for us to appreciate what an advancement this was for people back then. The idea that you could hear people talking and playing music from someplace far away, and everyone within range of the radio station could all hear it simultaneously, was miraculous. Now, we take such things for granted. Radio broadcasting has had an impact on the public for 100 years.

The early equipment to generate a radio signal was crude. The radio receivers were, too. Both ends of the transmission were tricky to adjust, and difficult to make work well.

One of the big problems at the transmitter end was the fact that an AM transmitter has a maximum level that cannot be exceeded, somewhat like digital audio recording. That limitation was set by the laws of physics, at least for AM broadcasting. FM did not exist at that time.

Exceeding the maximum modulation level could result in equipment damage. At best, it might knock the station off the air momentarily – or for days, if the overload was prolonged and major parts had to be replaced. Even a slight amount of overmodulation would cause severe distortion.

Radio engineers had to monitor levels very carefully to prevent overmodulation. They had to be conservative, which prevented distortion and damage, but did not use the full potential of the transmitter. Good board operators were skilled at keeping the levels at the highest possible level, without going so high that it was dangerous.

The radio receivers had a different problem. In many communities, multiple stations could be received. And the nature of the frequencies used for AM broadcasting meant that at night, signals could travel thousands of miles.

The problem was that the signal strength varied considerably from one station to another. A nearby, powerful station could easily be 60dB louder than a weaker one. Tuning a radio back then meant keeping you hand on the volume control knob to prevent blasting your ears when a strong station was tuned in.

Early adopters of radio were technically-minded and could understand and adapt to this extreme range of loudness. But radio quickly became mainstream, and most people found the constant adjustment of the volume control to be unacceptable. They just wanted to find a program they liked, not fiddle around with the radio.

To address both problems, smart electronics designers designed a vacuum tube that could vary its gain as needed.

Without getting too technical, the concept was to provide a control voltage, superimposed on the audio, or in the case of the receiver, RF, coming into a vacuum tube amplifier grid. The grid is the element in a vacuum tube that controls a much larger voltage, which means the output is a higher level than the input. An amplifier, in other words.

Vacuum tubes are not particularly linear devices. That’s true of magnetic tape, too, and the answer was to “bias” the audio signal into the most linear portion of the tube’s capability. In a normal amplifier, the bias voltage is fixed, to provide the most linear response.

By the way, the same concept of bias is used in solid-state amplifier stages, for the same reason.

To achieve an “automatic volume control” function, the bias on the grid of the tube had to be varied, yet still remain within the linear range of the tube. This was not possible with most tubes, so a new approach had to be invented for this to work.

With careful construction of the grid, the tube could operate in its linear range while its overall gain could be varied with a change in the grid bias. It was an amazing invention, and we still use that concept in vacuum tube compressors and limiters.

The operating parameters of tubes are described in the technical specifications, which often use Greek letters to signify a particular characteristic. The one used to specify the gain of the tube is the Greek letter mu, which when written looks like an odd variant of the letter u.

The tubes that had this characteristic were called “variable mu tubes,” since their gain could be controlled by an external voltage.

The early limiters for broadcasting had a relatively small range of audio signal level they could control, perhaps 20dB, but it was plenty enough to prevent damage to the transmitter.

 The radio receiver problem was similar, but the range needed was much greater. That was not possible with one variable-mu tube, so the receivers used multiple variable-mu tubes in the radio frequency amplifier section. It was common to have at least three tubes that amplified the signal and whose gain was varied. That provided the 60dB of range needed.

 But where does the voltage that adjusts the gain come from?

It’s essentially the same in any compressor. The signal at or near the output of the compressor circuit is split between the output of the compressor and a separate amplifier section called the “sidechain.”

The sidechain amplifiers boost the signal to a fairly high voltage, perhaps 100 volts or so. It’s still audio at that point.

A rectifier is a device that converts alternating current into direct current. The early limiters used a type of tube called a diode to make the conversion. That provides a DC signal that varied with the level of the audio. It’s not quite that simple, but you get the idea.

If we take that varying audio voltage from the sidechain and rectify it, we can use that voltage to change the gain of a variable-mu vacuum tube. A level control, labeled “Threshold” is used to adjust how much compression is applied. The threshold control is just a gain control in the sidechain. That’s all there is to making a compressor, except for a few more details.

It’s those details that determine the characteristics of the compressor or limiter. By tweaking the properties of the sidechain output, we can adjust the attack time, release time, compression ratio, and several other characteristics. Those define the sound of the compressor.

All this can be boiled down to a bunch of mathematical equations, but the real goal should be to make something that controls the levels in a way that sounds natural. The sound of the music, or a voice, can be impacted severely by these parameters, and picking the best points for the esthetic goals is the challenge for the designer.

Our hearing has built-in limiting. The threshold is more or less fixed at a sound pressure level that starts to become injurious to our hearing. I knew about that, and during construction of my first studio, I really became aware of the inherent limiting in our hearing, and got a sense of its characteristics.

During construction, one worker was busy sawing lengths of wood on a table saw, while another worker was hammering in nails. I quickly realized that the steady roar of the table saw was modulated by the hammering. Each hammer hit dimmed the level of the saw for a short time. It was like putting compression on music with a super-loud kick drum -- all the other instruments and voices are modulated in level by the kick.

What was fascinating to me was the attack and release times of our built-in limiting. I realized that those time constants were very similar to what I found pleasing when adjusting a compressor. It sounded like a moderately fast attack with a release time of less than a half second.

I wondered if that is why I prefer those time constants in the recording process.

I also wonder if everyone has the same characteristics. Or is it an individual thing? I doubt that anyone has ever researched this. For one thing, it would require exposing people to sound pressure levels that are way beyond the OSHA limits.

 The variable-mu compressor was the first type invented and continues to be popular with many recording engineers and producers. It has a unique and appealing effect on the music. Maybe because it emulates the “natural” limiting in our hearing.

The compression is imperfect, from a purely technical point of view. There are dynamic changes that occur during the compression process that result in a hard-to-define defects in the sound of the compression.

One that I can hear and measure is a certain amount of overshoot in the limiting attack with many compressors. The level is initially reduced significantly and then quickly rebounds to the set threshold level. This happens very fast – within milliseconds or less. But we are able to hear this happening. Listen for it the next time you use a variable-mu compressor. This characteristic is also audible in some other compressors. For the most part, we find this sound acceptable and it is part of the classic compression sound we like.

 Distortion increases with the amount of compression. With well-designed tube compressors, the distortion tends to enhance the music, mostly from an emphasis on the second harmonic. That’s an octave above the original note and our ears find that pleasing. It’s musical. It’s the sound of many of the instruments we record.

Interestingly, the tube most used in the classic tube compressor, the 6386, was developed from the old radio frequency amplifier tubes used in radio receivers. Various generations of that design were developed over the years, generally becoming smaller. But the basic sound of those tubes has not really changed over the past 90 years.

 Broadcast engineers in the 1930s realized that the compression amplifiers allowed them to run higher levels of program material. That resulted in a radio broadcast that was noticeably louder than it would be without compression. That was a benefit for AM broadcasting since the medium is very susceptible to noise, both natural and manmade. By keeping the level higher, the listener was subjected to less noise. A noise-free signal could be heard at greater distances, too.

Of course, if a little of a good thing is beneficial, then more must be better still. That led to radio people wanting more compression. To this day, radio station managers are obsessed with having their station be the loudest one on the dial. That loudness war has become ridiculous, and I believe it is a major factor in the decline of radio listenership.

All the vacuum tube variable-mu compressors of that era, and today, use the same basic circuit. They differ only in the time constants chosen in the sidechain, and the quality of components used.

By the way, the automatic volume control in the radio receiver does not increase the apparent loudness. It simply reduced the differences in the strength of the received stations. No compression is applied to the audio in a radio receiver.

Variable-mu compressors were the only type available for at least 25 years. But then other ways of providing compression were invented.

 The “optical” compressor uses a light source originally developed for nightlights and other applications that just needed a small amount of light. Those electroluminescent panels provide an odd greenish light, and they produce practically no heat. They were designed to operate directly from the AC mains – about 120V or 220V, depending on the country. And they worked with alternating current – 50 or 60Hz.

The brightness of the light could be varied by the voltage output of the sidechain amplifier. If that light was directed onto a light-dependent resistor, the resistance of that special resistor could be varied over a fairly wide range.

Light-dependent resistors have many applications, such as automatically turning on a streetlight when the ambient light level gets low enough. Or automatically adjusting the light level of things like displays, depending on how bright it is in the room (or in your car).

If that light-dependent resistor varies the input signal to the compressor, the gain will be varied, just like in the variable-mu compressor. This approach has the advantage that the tubes are always operating in their optimum range, so distortion goes up less with more compression than it would otherwise.

The LA-2 was the breakthrough design, and it remains a great way to apply compression. That method of compression has a unique sound that often is the perfect complement to the music.

On the downside, the light-producing element changes its characteristic over time, and will eventually fail. There are hundreds of different electroluminescent light panels, and they all have different characteristics; modern replacements may not sound the same.

And the thousands of different light-dependent resistor types have similar problems.

It is difficult to exactly duplicate the original sound, but with careful selection of parts, those compressors can remain in service. They may not sound exactly the same, but they still provide a sound that is very useful.

 The transistor was invented by Bell Labs in 1947. It took a few years for the technology to stabilize enough, and for prices to come down enough, for transistors to appear in audio devices. The early transistor circuits did not sound very good at all, with a narrow linear range where they sounded acceptable. Beyond that range, the distortion rose. But as manufacturers improved the specs, and equipment designers learned how best to use the new devices, the quality of transistorized audio gear improved.

In a general way, a transistor parallels the physics behind a vacuum tube. However, transistors do not work well varying the bias to change the gain.

For many of us, the sound of transistors was never good for music. In an early episode in this podcast, I talk about why tubes have always sounded better to me.

It was the invention of a variant of the transistor, the field-effect transistor, or FET, in 1953 that a solid-state device actually shared more of the characteristics of vacuum tubes. No surprise that it was Bell Labs that invented the FET.

Without going into the technical details, the FET was capable of being a variable-gain device, somewhat like a variable-mu vacuum tube.

By translating the vacuum tube variable-mu circuit to one that used FETs and other transistors, an effective solid-state version could be made. It never sounded the same as the tube version, but it became a popular device. The 1176 is the best-known example of this.


By the 1960s, solid state device manufacturers developed a way to include many individual transistors in a single, small package. The IC, or integrated circuit, was born. This allowed much more complex solid-state equipment to be practical, especially digital circuits.

Analog IC devices were also possible. The solid-state operational amplifier, or op amp, was developed. The original op amp design goes back to the 1930s and used tubes, of course. The application was mainly for analog computers, the predecessor of our modern digital computers.

Instead of multiple tubes, the solid-state op amp has dozens of individual transistors in a single, small package. The op amp made large recording consoles possible.

Along those same lines, specialized ICs called voltage-controlled amplifiers, or VCAs, were invented. It was immediately apparent that the VCA could also be the control element in a compressor. Those compressors had their own sound. The dBx 160 is an early example.

There were a few other approaches to compression along the way, but the variable-mu tube, the optical panel/light-dependent resistor, and the VCA were the dominant ones for a couple of decades.

One unusual method of compression uses a pulse-width modulator circuit. This is the method I chose for the D.W. Fearn VT-7 Compressor. Others have used the same approach, which actually goes back about 100 years. Back then, it was mostly for things like controlling the speed of motors. That motor speed application using a pulse-width modulator is still used today.

The PWM compressor has the advantage of essentially no change in the audio performance as the amount of compression increases. The amount of distortion can remain very low, even with extreme compression.

A PWM compressor can easily provide 60dB or more of clean compression, although that much would rarely be useful. You can learn more about PWM compressors in an early episode of this podcast, about the VT-7 design.

 When computers became powerful enough for real-time audio processing, we had the invention of the digital audio workstation. And it became possible to implement the functions of a compressor or limiter in software.

Although many engineers prefer the real hardware, the “plug-in” software version of compressors and many other sound-modifying devices became a significant portion of the processing used in recording. It works for all kinds of emulations, not just compression.

The translation from a hardware device to a software equivalent is tricky, but the designers get better all the time with their emulations.

One thing you can do in software is to design a compressor from scratch, with just the characteristics you want. You can eliminate all the “defects” in hardware compressors, for better or worse.

But the best part of a software limiter is that it can “look ahead” at the incoming audio stream and essentially have instantaneous attack time. This is especially useful as a safety limiter in a digital recording system. All hardware compressors and limiters have a finite attack time. It can be quick, but never instantaneous. An instantaneous limiter is perfect for keeping levels as high as possible without exceeding digital zero.

 Back in the early 1970s, I often speculated about an “anticipatory limiter,” that would already know what was coming and its gain reduction could be implemented just an instant before the peak actually occurred. It was mostly a joke, invoked when the existed limiters were not quite fast enough for a particular application.

Soon, the first digital delay devices became practical. I was an early adopter. They didn’t sound very good, with their low sample rates and limited bit depth. They were OK for providing the equivalent of a tape echo. They had the advantage of being adjustable over a wide delay range.

Soon after acquiring one of those digital delays, I was anxious to try my anticipatory limiter idea.

It worked, but not very well. For one thing, the minimal delay through the box was too long. The A to D and D to A converters had significant latency, so the minimal delay time was more than 10 milliseconds. What I wanted was a delay of less than a millisecond. But I decided to try it anyway.

I ran the audio track into the digital delay and then to the limiter, but the un-delayed audio also went into the limiter control input. That way, the limiter acted before the audio got to it.

It did not sound good at all. For one thing, the original audio was degraded by the low-quality delay unit. And, of course, the delay was too long. The result was an interesting effect, but I could think of few applications where that sound would be useful.

The concept was sound, however, and when the first digital limiters appeared, I was interested in hearing what they did.

I was pleased by the almost complete transparency of the limiting, at least at moderate levels of gain reduction.

I was never one to strive for maximum loudness, but the digital limiter certainly helps with loudness.

In my work, I often have a digital limiter as the last thing in the processing chain. The way I use it, it never does very much. It just catches those peaks that would otherwise bring the level of a song down a dB or two.


So, is there a purpose for a compressor, beyond making things louder or preventing a peak from ruining a digital recording?

I use compression in a variety of situations to improve the final product. Many instruments have different loudness characteristics. An electric guitar, for example, into a tube guitar amp, has a very limited dynamic range. In the right hands, that can still be a powerful sound.

But the electric guitar sound may be too dense when other instruments with a wider dynamic range are playing. You can make It work by balancing the guitar and the other sounds to achieve a better mix, but often that is not the best solution. I use the electric guitar as an extreme example, but this concept applies to many instruments – even a harpsichord, which has only one level, no matter how hard you hit the keys. That’s an extreme example, but many instruments have some degree of what is essentially built-in maximum level. The result is a very dense sound.

You might want to make sure you can hear all the words on a vocal track. It may be desirable to put some compression on the vocal. That gives it a fighting chance against other instruments who might be much denser, with consistently high loudness.

The singer might want to vary their singing level according to how they want to convey the emotion of the lyrics. Some words may get lost. Other singers may run out of breath before the end of a phrase and their level drops. Others have trouble maintaining a constant distance from the microphone. Compression can minimize these differences. And that keeps the overall energy of the song up.

You can often achieve the same thing, with much finer control, by automating the level of the vocal where needed. That often makes the vocal performance more impactful, without being too obvious. And a combination of compression and level automation may be a solution, too.

 There are other instruments, like an electric bass guitar, that often benefit from compression, even though the level between notes may be pretty consistent. The characteristics of the compression may fit the bass perfectly, to improve its punch.

Beyond those examples are many other sounds that might benefit from compression. But I feel it is a bad idea to compress all the tracks individually, at least for many types of music. That can result in a song that is fatiguing to listen to. That might be a good solution if you want to annoy the listener, which can be a legitimate goal of a recording. But most of the time, I think it detracts from the impact.

A better solution, in my experience, is to add a bit of compression to the overall mix. It usually doesn’t take much. I routinely have a VT-7 compressor on the mix bus with just a dB or two of compression. It doesn’t seem like much, but it can make a big difference in the overall impact of the song.

It all depends on the music, of course. Whatever we do should always be in the service of the music.

After the compressor, I usually have a plug-in peak limiter. I set it so that it doesn’t do anything most of the time. It just pulls down the occasional peak that would otherwise require an overall reduction of the level of the mix to avoid digital overload.

I find that when I do that final limiting, I set the output ceiling level to -2dB, referenced to digital zero. That just sounds better once the song gets to the listener. You will find that the true peak level, interpolated between peaks will then be about -1dB. It has been my observation that most consumer equipment starts to add a lot of distortion on levels above about -1 or -2. The loss in loudness is not severe, and the listener gets a better product.

That works for me because most of what I record is acoustic music. And I am really annoyed by distortion – at least the nasty-sounding kind. But I have found this approach also works well with high-energy music like punk or dance music. Most listeners experience less fatigue and more joy when what they hear is very clean.

I admit that I am not in the mainstream with my approach, so you have to decide what is acceptable to you.

Most compressors and limiters have a variety of controls that you can adjust. Some are very basic, like setting the threshold. That simply adjusts the amount of compression. The make-up gain control restores the output level of the compressor to provide the proper level for whatever is next in the audio chain.

As a rough rule, the amount of makeup gain required is about equal to the amount of gain reduction. If your compressor has a meter, or string of lights to indicate the amount of compression, that gives you a good idea of the amount of compression. But like most things in audio, it isn’t really that simple. A mechanical meter cannot follow the peak level of a signal, but with experience, you can estimate the peak level based on how the meter responds. This varies from compressor to compressor, since many of them do not use a standardized VU meter for this function. That’s OK because you should be setting the amount of compression based on how it sounds, not on what a meter shows you. However, a decent meter can help get you in the ballpark.

Same with the output level. A VU meter shows you a kind of average level. It corresponds very well to our concept of subjective loudness, but it isn’t precise.

And just because a meter has a face marked like a VU meter does not mean that it is a true VU. The specifications for a VU meter are quite extensive and have a lot to do with the ballistics of the moving meter needle. You will have to get some experience with your compressor to know what the meter is telling you.

A string of LED lights or segments show basically the same thing, but that type of meter is capable of showing peak levels much more accurately. If you feed that same signal into different compressors, one with a random meter type, one with a true VU meter, and one with a peak-reading LED meter, they will all show different amounts of compression.

It may not matter much in the real world, but you should be aware that different meters will present the information to you in different ways.

So, how much compression is good? There is no answer to that. It depends on how it sounds. But if I were to make a general statement, I would say that most engineers use more compression than is optimum.

What about those controls labeled Attack and Release? Well, in the simplest terms, the attack adjusts how quickly the compressor responds to a level peak. A very fast attack time will capture and bring down the level of the most intense peaks. These peaks usually occur at the start of a percussive note. We sometimes call them transients, due to their short duration.

Do you want to round off that percussive attack of, say, a snare hit? Maybe, maybe not. I think it is better to allow that transient peak to get through, and then the compression can take place. That gives you a much more dramatic version of the snare sound. The snare will sound much less explosive with a fast attack time. You have to adjust the attack time by ear.

I use a snare drum as an example because the effect of compression is very easy to hear. But the same applies to lots of sounds that musical instruments make. There is a reason why a piano is classified as a percussion instrument.

Other sounds may be able to tolerate a faster attack time without compromising the sound.

 If the compression is on a sub-mix or the entire mix, the calculation changes. You have to find an attack time that works best for the overall composite of all the sounds. That sounds like a compromise, but it may work better than you might first think. There is something about overall compression occurring on all the sounds in the mix that is different from compressing each track individually. To my ear, the mix sounds more cohesive with overall compression.

 The release control adjusts the amount of time it takes for the gain to come back up after it is dropped down by the compressor. A fast release time makes the music sound very intense and kind of desperate, which may be exactly what you want. You have to decide.

A longer release time smooths out the levels and provides a more relaxing effect on the mix.

I tend to use a moderately slow attack time and a medium-long release time. But it always depends on the music.

Note that on many compressors, you can set the attack and release times so short that the compressor will respond to the individual cycles of notes below a certain frequency. This is rarely possible on most musical instruments, but it can do strange and unpleasant things to bass instruments. If you hear this happening, making the attack and/or release times just slightly longer may clean up the bass. Or maybe you like that sound, which can be effective with some music.

 One common control adjusts the “ratio.” This simply indicates how much the output will increase for a given amount of input. A ratio of 2 to 1, for example, means an increase in level of 2dB going in becomes 1dB coming out.

A 10 to 1 ratio means that when the input goes up 10dB, the output only goes up by 1dB. That’s pretty severe, but it is useful sometimes. Some compressors allow even higher ratios.

When the ratio is 10 to 1 or more, common practice is to call the device a limiter instead of a compressor. There is no real defined point between compression and limiting, and many people use the terms interchangeably.

The ultimate ratio is the brick wall digital compressor, which does not allow any peak to exceed the set level. That generally does not sound very natural, but it works well to protect things like a digital recording, or a broadcast transmitter from ever exceeding the maximum permissible level.

 The transition for linear operation, meaning no compression, and the onset of compression can be gradual or abrupt. If you graph that function, it sort of looks like a human knee, and the term “knee” is often used to describe this characteristic. A sharp or hard knee has a rapid clamp-down of the level, while a soft knee is much more gradual and less obvious. The shape of the knee should correspond with the purpose of the compressor. Some compressors allow you to adjust the knee. Use your ears to hear what this does.

 Many compressors have a switch to insert a high-pass filter into the sidechain. This rolls off the low frequencies going into the sidechain, which means there will be less compression on the lows than on the mid and upper frequencies.

Why is that useful? Well, with bass-heavy recordings, the kick drum and bass may take over the compressor and it will respond predominantly to those instruments. That can work well for some type of music, but it can also “punch holes” in the mix on every kick drum hit. The sidechain high-pass filter is another thing that requires your judgment to set. Using the sidechain high-pass filter increases the level of the low frequencies in the mix. Without the filter, the sound may become bass-deficient.

On a stereo bus, it is necessary to compress both channels equally. Otherwise, the stereo image will tend to shift left and right depending on the level in each channel. Sometimes that does not matter, but usually it does.

So, compressors often have a way to link the two channels, so that the compression on both channels is always the same. That keeps the stereo image stable and it is the way most people use a mix bus compressor. As always, you can try it both ways and see which you prefer.

 Other controls may be available to adjust additional parameters. I won’t go into the wide variety of things that some compressors let you modify. You will have to experiment and see how the sound is changed.

Some compressors use multiple gain reduction stages, each designed to only affect a single band of frequencies. That allows the audio to achieve even higher loudness, since you can optimally set each band for the attack and release times, and the amount of compression. By adjusting the relative level of each band on the multi-band compressor, you can shape the overall frequency response, much like an equalizer.

Personally, the multi-band type of compressor does not usually fit my recording style very well, but it might be a good solution for you.

 And a final note on loudness. Although everyone initially likes a very loud recording, it is ironic that too much compression may actually make the music sound smaller and less exciting. Most music thrives on its inherent dynamic range. When we reduce that range too much, the song suddenly sounds boring.

Too much compression may cause the listener to be fatigued by the music. This is an interesting psychoacoustical phenomenon. Not much research has been done on this topic, but experienced engineers can foresee how a fatiguing mix can make it difficult for someone to listen to. Real music utilizes loud and soft sections to add drama and convey the emotion of the story. Excessively limiting can eliminate that valuable trait of the music.

 Compression and limiting are huge topics and I have only covered the basic things in this episode. But that is probably enough to give you more insight into what your compressor is doing and how to adjust it.

I believe that the more you understand the workings of your gear, the better you can utilize it. But the music always comes first. You have many tools to modify the sound, presumably to make a more effective recording. Whatever you do, make sure the music is improved by what you are doing to it. And the only way to know that is to try to experience it as a typical listener.


I am always pleased to hear your comments, or suggestions for future episodes. Reach me by email at

 This is My Take on Music Recording. I’m Doug Fearn. See you next time.