How to choose the right audio preamp to match your music, instruments, and budget.
The Golden Age Project PRE-73 MKII has circuitry based on classic preamps to deliver vintage sound that’s both noise-free and has a wide dynamic range.
If you want to know what a preamp is and what one can do for you, read on. You’ll find answers to all your questions about how preamps works, the various types that are available, and which preamp will be best for you, your budget, and your music.
We’ll first address preamps designed to handle signals from mics, then explore instrument-specific preamps.
Table of Contents
What is a preamp? What does it do?
Types of Preamps
What Is a Channel Strip?
Transformers and Their Effect on Preamp Sound
Which preamp is right for me?
So, which preamp do I buy?
Still have questions? We have answers
Glossary of Preamp Terms
What is a preamp? What does it do?
A preamp’s job is to take an audio signal, whether from an instrument or microphone, and give that signal sonic shape and definition before the signal is amplified for further audio processing or output through speakers, headphones, or other audio gear. The preamp’s role is crucial in creating quality sound.
While this may sound like a very dry, functional step in the signal chain, a preamp can add immense character, warmth, and dimension to your music.
I already have a mixer and/or digital interface. Do I still need a preamp?
If you already have a hardware- or software-based recording rig, you might wonder how essential an external preamp is. The truth is that, while there are exceptional mixers and interfaces on the market today, there is usually room for improvement in the quality of the sound this gear delivers.
Your audio interface or mixer will in all likelihood offer some preamp functions. While a DAW (Digital Audio Workstation)—the platform that allows you to record and mix your audio in a digital format—excels at letting you combine tracks, add virtual instruments, and generally manipulate your audio material, it is usually limited in its ability to deliver professional-level sound quality by the level of quality of the equipment feeding audio into and out of it. Realistically, there isn’t room in a single interface or mixer to house the components of a well-designed external preamp. It’s also unreasonable to expect a moderately-priced interface or mixer to deliver the sound quality of a dedicated hardware preamp.
The beauty of adding a high-quality preamp or channel strip to your setup is that instead of paying $4,000 per channel times 32 for a pro-quality mixing console, about $2,000 to $3,000 buys you a “gold channel” for your DAW or mixer that you can use to record track by track. That’s one of the secrets world-class sound on a project studio budget. If you need to record more simultaneous channels, there are a number of high-quality multichannel preamps available from PreSonus, Grace Designs, Millennia, and others.
Even if your budget doesn’t permit buying a single, high-end preamp, there are many moderately priced models that can still deliver excellent sound for capturing your bedroom concertos or garage band jam sessions.
If you’re operating with a really tight budget, there are lots of interfaces available that include preamp functions that may serve your needs. To explore all the options, read our Audio Interface Buying Guide.
Types of Preamps
The first important distinction to be made among the different types of preamps available is an aesthetic one: whether they’re designed to add color or deliver a transparent signal. You’ll usually find both types in most studios to meet a range of recording needs. A color preamp can add body or character to a thin voice or instrument, whereas a transparent preamp strives to reproduce the original instrument or vocal sound as accurately as possible.
Preamps That Add Color
Some preamps are designed to enrich the overall sound or timbre of your signal. This can be useful if your signal sounds dry and thin. Adding color with a preamp can make your recordings sound intimate or over the top and larger than life depending on how you apply that color. Audio engineers who prefer color in their preamps subscribe to the philosophy that whatever helps you achieve the sound you are after should be employed in the process. Generally, those in this camp prefer tube and solid state preamps with transformer-coupled designs.
However, there are other audiophiles and engineers who subscribe to a more pure, transparent philosophy of sound. They would argue that the goal is to reproduce the sound of the instruments and voices as clearly and truly as possible without altering their fundamental tone or timbre along the way. These producers and engineers prefer to use preamps that provide as transparent a signal reproduction as possible. They tend to prefer transformer-less solid state preamps. Recording engineers who work with jazz and classical performances that are focused around acoustic instruments often seek a more transparent signal.
Though some producers and engineers prefer transparency while others rely on preamps to be a part of the fundamental sound, there is no clear right and wrong here. Transparent does not mean sterile. And “color” doesn’t mean doing away with the sounds and instruments and voices in your arsenal. Both philosophies can be applied to recorded music to achieve great results. In fact, many producers and engineers prefer to use both types of preamps, selecting one or the other as the situation and music dictate.
When considering how preamps affect your sound, it’s important to consider that any electronic signal path will always add, change, or take away something in the process. The only true acoustic reproduction is done by your ears. Once you factor in the impact of microphones, cables, preamps, equalizers, compressors, analog to digital converters, and loudspeakers, the end signal is going to be affected. The ultimate question is what sounds pleasing to your ears.
Preamp types can be also be categorized by their circuit design. The first main distinction is tube versus solid-state preamps. When you hear or read discussions regarding preamps and their sound, this distinction will usually be at the forefront.
Tube preamps use thermionic tubes (also called vacuum tubes, or among our British friends, valves) to create gain. Most tube amps will color your sound quite significantly. Tube preamps are known for adding deep bass, open and airy highs, and warm presence in the midrange. Often, tube preamps are described as “warm”- or “fat”- sounding. To understand what this means, let’s look at how tubes function.
As the signal level increases, the tube produces mild distortion, which in turn produces the body and depth that “warmth” is meant to describe. Since the distortion is produced gradually and in measured amounts, it is pleasing to the ear.
Going deeper, the distortion takes the form of second- or even-order harmonics, which are considered to be more “musical.” Here’s why: Harmonics are component pitches of a fundamental tone that sound at whole number multiples of the fundamental. Musically speaking, the second-order harmonic forms the interval of a perfect octave. The octave reinforces the fundamental tone, similar to the effect known as doubling. That’s where the word “fat” comes in since octave doubling results in a broader sound with more depth. “Beefier” is another word used to describe tube preamps by engineers and producers.
Even-order harmonics produce pleasing distortion (much like power chords on guitar) since they don’t produce the dissonant pulsing (intermodulation) caused by odd-order harmonics. The addition of even-order distortion via tubes equals color. This effect is often referred to as harmonic distortion.
Tube preamps also provide color through the natural compression characteristics provided by tube circuit designs. Looking at waveforms, the curve at the top of the wave becomes more compressed as gain is applied gradually (or when a tube is overdriven). The human ear hears this natural compression as pleasing. Natural tube compression can be a subtle effect, but many engineers and producers will pay top dollar for high quality tube compression on their final mix. It is the “audio glue” that brings a mix together and gives it a professional, complete sound.
Yet another natural function of tubes is to act as low-pass filters, which smoothes out distortion in the high end of the spectrum and can somewhat reduce high-frequency content. This can be desirable or undesirable depending on your sonic goals. For example it’s desirable for warmer vocals and undesirable for percussion, where attack information is essential.
The Manley 40MP has selectable gain and is optimized to produce wonderful results with condenser and ribbon mics.
Browse the complete Musician’s Friend selection of tube preamps here.
Solid State Preamps
One result of evolving technology is the development of transistors to achieve gain in solid state preamps. Transistors are more efficient in this regard, creating gain with less heat. Transistors operate more consistently as gain increases, maintaining very low distortion up to their maximum levels (beyond which they exhibit extreme distortion or “clipping”.) They perform optimally and with minimal distortion at high levels of gain. The transparency of solid-state devices is largely due to their capability of accepting significantly higher gain levels without distorting.
Where tube amps have their own fan base centered around their unique musicality, the solid-state world has its legends as well. Known for their unique sonic signature are the solid-state preamps designed by Rupert Neve, who is widely considered the grandfather of professional audio. Neve’s designs can be found in some of the most famous recording studios around the globe.
The Warm Audio Tone Beast Microphone Preamp uses two types of op amps to produce a versatile range of tones from vintage to modern.
Browse the entire Musician's Friend assortment of solid state preamps here.
Another category is the hybrid preamp. Hybrids are sometimes considered a marriage of the best elements of both tube and solid state design and components. Hybrids typically use solid state circuitry in the input stage for maximum clarity without distortion, and output gain control that drives a tube, achieving the rich harmonics from tube saturation with which to season the overall sound.
The Summit Audio 2BA-221 features separate tube and solid-state outputs and handles both line and mic signals for great versatility in the project studio.
These preamps use solid-state circuitry like the hybrids above. However, instead of using tubes to add warmth, the signal is altered digitally to add warmth and other desirable sonic qualities. Combining digital processing with actual physical variations in the analog circuitry, modeling preamps can mimic the sound of virtually any preamp or compressor—tube, solid-state, or otherwise. And in any combination. Some purists will argue that digital processing cannot substitute for the real thing, but modeling preamps have earned a prominent spot in many of today’s top recording studios due to their versatility and reliability, and their ability to deliver this quality at relatively low prices.
The Focusrite Liquid Saffire 56 houses a total of 8 preamps, offers 10 different emulations of classic preamps, and 28 inputs and outputs for enormous connection flexibility.
With some similarities to modeling preamps, digital preamps convert an analog signal to a digital signal and add their sonic signature in the processing before the signal is output to a DAW. This allows the user to process their signal with a dedicated preamp and bypass the built-in process of conversion supplied by the audio interface’s typically inferior converters. Some digital preamps offer a digital output sound card that can be installed in the user’s computer. Others have the processing and conversion built into the preamp itself, such as those offered from Apogee.
Technically, these preamps could be considered digital interfaces, as they effectively convert an analog signal to a digital signal for use in a DAW. However, the units are designed mainly as preamps, and the digital conversion of the audio signal is a convenient byproduct of the primary preamp function.
The Apogee Duet features two professional-quality digital microphone preamps as well as analog-to-digital conversion for simple recording directly to most Mac or iOS devices.
Generally, the preamp principles we’ve discussed so far are based on using a microphone as the first component in capturing sound. Most preamps that fall under the categories we’ve covered are designed to take an input signal from a microphone and process that signal internally.
However, there is another category of preamp that is important, while also significantly different from the types discussed above. There are many different types of instrument-specific preamps available today. These preamps offer circuitry designed to optimize the tone of a specific type of instrument, such as electric guitar, acoustic guitar, or bass. Many of these units are classified as DI (Direct Injection) boxes that allow connection of guitars and basses directly to mixers and recorders.
You can learn more about direction injection with our Direct Box Buying Guide.
The A Designs REDDI Tube is in essence a bass guitar-specific tube preamp designed to optimize the instrument’s range and deliver unparalleled bass tone.
Instrument-specific preamps are designed to take into account the unique impedance requirements and voicing of the instruments they’re designed for. In many cases, instrument-specific effects will also be included in this type of preamp.
You can browse a selection of electric guitar-specific preamps here.
The Tech 21 Sansamp Bass Driver combines preamp, stompbox, and DI functions, delivering a quality bass signal wherever you need it.
What Is a Channel Strip?
Basically, you can consider any preamp with additional signal processing circuits to be a channel strip. The channel strip of today is a standalone unit that in its most common form includes a preamp, equalizer, and compressor/limiter. Some channel strips offer variations on that theme such as preamp and EQ only, or preamp and compressor only, while others offer additional sections such as a de-esser or enhancer. In short, a channel strip can offer all the signal processing tools you need in one box.
Initially the console channel strip grew out of the need for convenience. When most recording was done in the major studios, engineers didn’t want to spend the extra time connecting various units together. Mixing console channels thus were given additional capabilities beyond just microphone preamps, including EQ, and in some cases, compressor/limiters. Modern day channel strips are often similar to a single channel from a large, traditional mixing console— just taken out of the large console frame and placed into a smaller, rackmountable enclosure.
Typically, channel strips tend to employ hybrid or solid state designs.
The Universal Audio LA-610 Mk II Recording Channel is an updated version of Bill Putnam’s channel strip that captured the sound of many of the world’s most revered recordings.
Learn more about how channel strips evolved with our article History of the Channel Strip.
Browse the full channel strip collection at Musician’s Friend here.
Transformers and Their Effect on Preamp Sound
One of the main factors that determines color versus transparency in preamps lies in the use of transformers. Much like the PC/Mac, automatic/manual transmission, or dog/cat debates, audiophiles are split between the use of transformer-based circuitry to provide color and transformerless circuitry that delivers the purest, most transparent signal.
As part of a signal path, a transformer will always add some level of color to a signal. Reducing the number of active “colorful” components in a signal path is arguably the way to achieve more audio transparency. Most often, you will find preamps without transformers to be of the solid-state variety (like the Grace Design M101).
Transformerless tube preamps exist (like the Millenium M-2B), but are much harder to find and tend to be expensive due to the difficulty in building them. George Massenburg and John La Grou are two of the more well-known proponents of transformerless design.
The Grace Design M101 Mic Preamp is intended to produce a transparent signal and has no transformers in its circuitry.
The main function of a preamp is to take a low-level input signal and increase it to line level. A transformer’s function is to act as a bridge between the input and output sections of a preamp while increasing signal voltage. The transformer does this through magnetic coupling and induction. Transformers offer the advantages of better circuit protection and true isolation due to the lack of hard wiring between input and output. Quality solid-state transformers are considered desirable for their stability and transparency along with the ability to add a unique color to the overall sound. The defining factor in transformer-based preamps is the quality of the transformer itself. This is the reason that many manufacturers have their transformers custom made to precise specifications.
Which preamp is right for me?
If you’ve already read through the above categories about preamp design, you should have a basic understanding of how design affects the overall signal path. Choosing the right preamp for your project comes down to deciding what you want your preamp to do for you. Ultimately, your ears and your budget should be your guide.
If you are looking to add warmth or body to vocals, you should consider a tube preamp that can add some of the “beefy” color you need to fill out your sound. One popular yet inexpensive option is the PreSonus TubePre V2. If you’re on a tight budget, this preamp can add a lot of color to your sonic landscape for minimal cost. If you have a larger budget and want a world-class preamp, then the Universal Audio UA-S610 will provide you with the silky vintage warmth that you’re looking for.
If you want to record a rich-sounding instrument such as an acoustic guitar, you may not want to add any color, and instead capture the guitar’s sound as cleanly as possible. As guitars are often recorded in stereo, an excellent option would be the Millennia HV-3C. This award-winning stereo preamp uses high-quality components and an advanced design to deliver ultra-transparency to your audio.
Browse the complete selection of preamps at Musicians Friend here.
Additional Preamp Features and Functions
Many preamps come with multiple channels, allowing you to process and record multiple inputs simultaneously. Many affordable preamps, such as the PreSonus BlueTube DP V2 feature two channels. As you get into bigger projects, you may find you need even more inputs and there are preamps available for that.
Other common preamp features include gain / saturation control, EQ controls, compression, and more. Multiple I/O formats can also be a big advantage in integrating your preamp into your studio as it grows. In addition to the usual ¼” and XLR connections, some preamps offer USB, FireWire, S/PIDF, and other digital connectors making them DAW-friendly.
Owning Multiple Preamps
As you can see, different preamps have very different capabilities and sonic signatures. If you have a studio in which you may be doing an acoustic singer-songwriter session one week and a rock band the next, having a versatile selection of preamps is as important as having various types of mics.
As you gain experience and take on varied projects, you will find that you have the need for more than one type of preamp to deliver the best sound possible for each situation. Don’t worry, this is normal. Eventually, we all become gear junkies!
If owning multiple preamps isn’t an option with your budget, or you haven’t quite committed to the idea of owning multiple preamps, then consider something versatile like the A Designs MP-2A. With two separate channels you can record instruments in stereo, and the two tone switches offer four unique tone signatures for great plenty of tonal options.
The versatile A Designs MP-2A offers separate channels for stereo recording and tone switches that add character to your projects.
You can explore the huge range of preamp models at Musicians Friend here.
So, which preamp do I buy?
Determine how many inputs you will need, what kind of sound is most essential to your projects, what additional features you want, and go from there.
Learn more about how preamps fit into the overall recording process with our Recording Gear Buying Guide.
We want you to be pleased with your preamp purchase, and offer a 100% satisfaction guarantee and generous return policy so you can order your new preamp with confidence.
Still have questions? We have answers
After reading this guide, if you’re still not sure which preamp is right for you, we invite you call to one of our friendly and knowledgeable Gear Heads at (877) 880-5907.
Glossary of Preamp Terms
9-pin - Generic term for a 9-pin D-Sub connector mainly used for video synchronization. Alesis ADATs also use the 9-pin format for synchronization.
ADAT - An acronym for Alesis Digital Audio Tape, which is a digital tape recording system developed by the Alesis corporation in 1991. Known simply as ADATs, this series of modular digital multitrack recorders revolutionized home and professional digital recording.
ADAT Lightpipe - “Lightpipe” is jargon for the fiber optic cable used to inter-connect units equipped with ADAT optical ports.
ADAT Optical - A proprietary optical digital interface specification by the Alesis Corporation that allows the transmission of eight channels of digital audio through a single fiber optic cable at a maximum rate of 24-bit/48kHz. Higher sample rates such as 24-bit/96kHz can be transmitted via an ADAT optical interface, however, the number of available channels is reduced. (See, S/MUX)
AD/DA Conversion - An abbreviated way of saying “analog to digital and digital to analog conversion. (See Digital Converters)
AES/EBU - An acronym for the Audio Engineering Society/European Broadcasting Union). Officially known as AES3, AES/EBU is the name often used for a professional serial interface that transfers digital audio between devices. AES/EBU is not tied to any sampling rate or audio standard. Sometimes referred to as AES for short, the three-prong connector and jack are the same as those used for balanced XLR cables, however, the impedance for AES cables differs from that of mic cables. Standard analog audio cable impedance is 45 ohms to 70 ohms, whereas AES/EBU is 88 ohms to 132 ohms. To avoid digital errors due to impedance mismatching over such a broad tolerance, it is recommended by Belden CDT Inc. (makers of AES cables) to use 100-120 ohm shielded twisted pair AES cables.
Algorithm - A step-by-step problem-solving procedure, especially an established, recursive (repetitive) computational procedure for solving a problem in a finite number of steps. (American Heritage Dictionary)
Amplitude - How loud a signal is; the higher the wave goes above the median, the louder the signal will be. In an electronic system, amplitude is a measure of how many volts the signal is above the median voltage or average operating level.
Analog - In language we use analogy, which generally takes the form of a story, to describe a concept in a way that’s easier to understand. In audio, analog is a representation of physical waveforms using voltage or current. The fact the representation is continuous rather than stepped (varying between two discrete voltage levels) is what differentiates analog from digital information, and the reason why we tend to find analog signals more pleasing to the ear. After all, the real world is analog.
Attack - The initial portion of a sound, such as the pluck of a string or the snap of a snare, to the point at which it reaches its maximum volume, is considered the attack. The leading edge of the waveform at the peak of the attack and before the sustain portion of the sound begins (which occurs after the peak level drops somewhat) is called a transient. As such, transients tend to be the loudest part of the sound and often difficult to control. Attack on a compressor is the time it takes for gain to be reduced after the level of a signal passes the threshold.
Audio Interface - Much like a mixer, an audio interface will combine analog inputs and preamps. However, after the preamp, the signal is converted to digital information and sent via digital output to a computer. The digital information is processed via software and then returned to the interface where it is converted back to an analog signal for playback (see A/D/A conversion).
Balanced/Unbalanced - A method of wiring audio cables that serves to eliminate noise by virtue of common-mode rejection. The balanced cable consists of two twisted wires and a shield. One wire is considered positive with respect to ground, the other negative, and the shield, which carries no signal, is an outer wrapping of conductive material that is grounded at both ends. The signal is transmitted over one wire and received back on the other, while the shield does exactly as the name implies. In a properly balanced cable, impedance is equal on both lines relative to ground. This guarantees that noise is picked up equally on both wires. The balanced input of a preamp for example, only amplifies the difference between the lines, thus eliminating the noise that is common to both. In an unbalanced cable, there is only one wire that carries the positive signal, while the negative signal is carried on the shield. Any noise induced into the cable will make its way into the audio signal.
BNC (Bayonet Neill-Concelman) - A miniature bayonet locking connector for coaxial cable. The name BNC comes from its creators Paul Neill and Carl Concelman. While at Bell Labs, Neill developed a connector that became a U.S. Navy standard known as the N connector. Concelman developed a bayonet version of the N connector, which became known as the C connector. Together, they developed a miniature locking bayonet version of the C connector, which was named type BNC after both of them. In audio, 75 ohm coaxial cables with BNC connectors are used to transmit word clock signals for synchronization of digital devices.
Boost - A term applied to the task of equalization, where selected frequency levels are amplified or boosted (See also, Cut).
Clipping - When the loudest peaks of a waveform “bump” against the ceiling of an amplifier, the peaks are flattened and appear as though the tops of the waveform was clipped off—distortion ensues.
Common-Mode Rejection Ratio (CMRR) - In long cable runs, noise becomes an issue. Cables can pick up radio frequencies (RF) and other interference, which is why balanced cables are used for their noise-canceling ability. CMRR defines to what degree signals are canceled at the input of a balanced system.
Compression - A positive-going event (something with a value greater than zero and not a negative number) such as a sound wave that exerts force on a medium such as the atmosphere, which increases density and pressure within it. In audio, compression is the act of reducing the dynamic range of an audio signal via a compressor or other dynamics processor. (See Compressor)
Compressor - An audio device that reduces the dynamic range of a signal, and in so doing, increases its overall average level. In essence a compressor “lowers the bridge.” It narrows the distance between the softest and loudest signal. With a narrower dynamic range, the softer signals sound louder in relation to the louder signals, which is why we experience an increase in the overall average level. Compressors can more easily be thought of as devices that either increase or decrease levels around a set point. That set point is determined by the threshold control and the amount of increase or decrease of output relative to input is fixed by the ratio control. For example, with threshold set at 10dB and ratio set at 2:1, for every 2dB above the threshold of 10dB, the output will only increase by 1dB. If the signal is decreased by 2dB, the output will decrease by 1dB. In essence, the loud signal has been made quieter, while the softer signal has been made louder, since it wasn’t reduced as much. The overall effect of reducing dynamic range is to prevent program material such as a vocal track that has wide leaps in level from overloading a recorder’s input. It also allows the track to “sit better in the mix” by preventing the loudest passages from standing out too much (“sitting on top of the mix”) and the softest passages from being buried in the mix. Due to its ability to effect timbral changes by selectively compressing specific frequencies and waveforms, compression is also used as a creative tool to fatten drum sounds, to increase bass sustain, to smooth vocals, and to bring specific sounds forward in a mix.
Convolution - A mathematical process of correlation between two functions. A convolution is an integral that expresses the amount of overlap of one function as it is shifted over the other. It therefore "blends" one function with another, and is the product of those two functions. In audio, convolution enables you to take the sound of a flute for example, and process it through a crystal vase. It is therefore a very powerful tool for creating realistic emulations of the sonic response of physical space or other hardware devices.
Cut - A term applied to the task of equalization, where selected frequency levels are attenuated (reduced in level). (See also Boost).
D-Sub (D-Subminiature) - D-Sub is a prefix used to describe multi-pin connectors commonly used in audio equipment. D-Sub connectors come in 9-, 15-, 25-, 37-, and 50-pin configurations. (See Cable Buying Guide)
DAW - An acronym for Digital Audio Workstation. In audio, a DAW can be considered an all-in-one hardware system that comprises a mixer, digital recorder, and sound-producing modules with editing features that allow you to create, record, and mix music. DAW also applies to hybrid systems that combine an audio interface, computer, and music creation/recording software. DAW can also refer to music creation/recording software alone.
dBm - A measure of signal power. The lower case m is an accepted abbreviation of mW for milliwatts. dBm is a unit of measurement expressing the relationship of decibels to power referenced to 1 millwatt. (See Decibel)
dBu - A unit of measurement expressing the relationship of decibels to voltage referenced to 0.775 volts. (See Decibel)
dBV - A unit of measurement expressing the relationship of decibels to voltage referenced to 1 volt. (See Decibel)
DC (Direct Current) - Functionally the opposite of alternating current (AC), DC does not change directions. While most components within an electronic device are DC powered, DC is not very healthy for audio signals, particularly those traveling to speakers, since DC produces no sound and uses a great deal more power.
DC-Polarized - Condenser microphones, which employ a capacitor sensor to convert atmospheric waves (sound) to electrical signals, commonly use a high DC voltage to polarize (positively or negatively charge) the plates of the capacitor sensor.
Decibel (dB) - Named after Alexander Graham Bell for its initial use in telephony, the bel is a unit of measurement used to indicate audio levels. A decibel equals one tenth of a bel. Since our ears are capable of detecting changes in sound in millionths of parts, the decibel, is logarithmic shorthand (based on powers of 10) that enables us to describe large numerical changes with smaller meaningful numbers. For example, instead of saying the dynamic range is 32,000 to 1, we say it is 90dB. In this example, which is a measurement of signal level, the value in dB equals 20 times the log of the ratio of two different signal levels. (Decibels are always a measure of ratio, which is one value in relation to another.) Sound pressure level (SPL), signal level, and change in signal level are all measured in dB. One dB is the smallest change in level that most people can hear, whereas 6-10dB is perceived by listeners to be twice as loud. Thus, when you see 0 dB on a meter, it’s actually a reference voltage. (See also dBm, dBV, and dBu)
De-esser - A type of compressor that responds to frequency rather than level. As the name implies, its primary function is to remove the “ssss” sound of the human voice when recording (called sibilance), however, a de-esser can also be used to tame harsh-sounding cymbals as well.
Desk - Industry jargon for mixers or recording consoles (used mainly by British engineers).
DI - An acronym for Direct Input or Direct Inject, a DI or DI box is also referred to as a direct box, which is a device that enables an instrument such as a guitar or bass to be connected directly to a mic- or line-level mixer input by converting the unbalanced high-impedance output of the instrument to a balanced low impedance.
Digital Converters - Generic name for processor chips that take analog information and convert it into digital information. Digital converters are divided into two categories: A/D converters, which take an analog signal and convert it to digital information, and D/A converters, which take digital information and convert it to an analog signal. In audio, it is most common to see the A/D converters in the input side of an audio interface or on the output side of a preamp (usually via an option card), while D/A converters are found on the output side of the audio interface. (See A/D/A Conversion)
Direct Recording - The process of recording an instrument such as an electric bass without a microphone. This is accomplished using a direct box to send a signal straight into a mixer, or a built-in hi-Z input on a preamp or digital workstation, etc. (See DI)
Distortion - Technically, any signal that passes through an audio system and is altered by it in any way other than volume. In a perfect system, the pure signal would remain unaltered. However, perfect systems don’t exist. Distortion, as it has come to be more commonly known, is either a severe alteration of a signal by virtue of unwanted components, artifacts, or noise; or a severe, yet desirable alteration (such as that produced by a guitar effects pedal). Ultimately, there are two types of distortion: the kind you like and the kind you don’t.
Doubling - Taken from the technique in orchestration whereby different instruments play the same phrase in order to create a thicker sonority. Doubling in pop music recording generally refers to a vocalist or guitarist performing the same part twice. Each take is recorded to separate tracks, which are then combined in mixdown. The subtle differences in timing and pitch create an interesting texture as well as a thicker sound. John Lennon of the Beatles was particularly noted for this technique. Often, the Beatles would combine a doubled lead vocal track with a single harmony track to create the illusion of a three-part harmony.
Drum Overheads - In studio jargon, drum overheads are microphones that are placed over the drum kit (about 3 to 4 feet above the drummer’s head) mainly to pick up cymbals. One mic can be used as a drum overhead, however they are usually a stereo pair of small-diaphragm condenser mics. (See Microphone Buying Guide)
Dynamic Range - In an audio unit such as a preamp, dynamic range is a measurement taken from the ratio of the loudest undistorted signal (output voltage) to that of the quietest signal (noise floor) expressed in decibels (dB).
DSP - Acronym for Digital Signal Processing. A powerful and flexible means of signal processing that employs algorithms and high-powered digital hardware (chips, computer CPUs etc.).
Electromagnetic Induction - Also called inductive coupling, refers to the transfer of a signal form one component to another through a shared magnetic field. A change in current flow through one induces current flow in the other. These components can be contained in a single unit, such as the primary and secondary sides of a transformer.
Enhancer - Rather than using EQ to bring out high-frequency content for more clarity, an enhancer circuit splits off harmonic content from a signal, amplifies it and recombines the amplified harmonics with the original signal. This allows a vocal or instrument to be brought forward in a mix without phase problems and accentuation of sibilant frequencies, which may occur by boosting high frequencies with an equalizer.
Envelope - Describes functions of an audio wave over time. The envelope of an acoustic wave comprises attack, sustain, and decay. Attack is the initial transition from silence to highest volume level, which is called the peak, sustain is the duration of the sound, and decay is the time it takes for the sound to fade away to zero. In synthesis, an envelope is made up of four components; attack, decay, sustain, and release (ADSR). It is still the function of sound over time, however, decay is the time it takes for the peak level of the attack to drop to the sustain level, while release (so called since it begins when a key on the synthesizer’s keyboard is released) is the time it takes for the sound to fade after the key is released.
EQ - Abbreviation for “equalizer.” Also used as a verb as in “to EQ” something.
Equalizer - Initially used in telephony to compensate for frequency loss during transmission, a device that makes the output signal equal to the input signal, hence the name, equalizer. The equalizer of today comprises a number of adjustable or fixed electronic filters that modify the frequency response (change in level over a range of frequencies) of an audio signal. You can think of an equalizer as a frequency-dependent amplifier. Equalizers are active or passive.
Even-Order - Harmonics Frequencies that are multiples of a fundamental tone by even integers such as 2, 4, 6, etc. A second-order harmonic would be the result of a fundamental multiplied by 2.
FireWire - Apple Inc. and Texas Instruments" joint implementation of the IEEE P1394 Serial Bus Standard. FireWire is a high-speed serial bus for peripheral devices that supports “plug and play” (automatic configuration) and “hot-plugging” (changing peripheral devices while running). Its first generation, FireWire 400, could transmit hundreds of channels of noise free, high-resolution digital audio and up to 256 channels of MIDI. Second generation FireWire 800 can handle twice as many simultaneous realtime streams.
Frequency - The number of complete cycles of a wave form in a given amount of time. Frequency is expressed in Hertz (Hz). One Hertz equals 1 cycle per second. As the number of cycles per second increases, so does pitch.
Frequency Range - The actual range or span of frequencies from low to high that a unit can pass or reproduce.
Frequency Response - A measure of the amplitude of a specific frequency at a given input level. As such, it shows amplitude variation (changes in volume) in output level across the frequency range. For example, a spec may show that based on a given input signal, the preamp will output 80dB at 20kHz, 78dB at 10kHz, and 83dB at 1kHz. The spec may be written as Frequency Response: ±3dB @ 40Hz – 20kHz. This means that within the given frequency range, output level will vary somewhere within a 6dB range [(-3dB) + (+3dB)] no lower or higher than 3dB. For signals produced beyond the specified range, the change in output level may be greater than 3dB.
Fundamental - The lowest tone or pitch of a harmonic series. Any object that can be set into vibration will produce a fundamental frequency and multiples of that root frequency called harmonics or overtones. The structure of the overtone series of a given vibrating body gives it a unique sound characteristic known as timbre.
Gain - The amount of amplification (voltage, current, or power) of an audio signal, usually expressed in decibels (dB).
Harmonics - A series of musical tones whose frequencies are multiples of a fundamental tone. Also referred to as overtones and partials.
Headroom - The dynamic range between the average operating level of a system and the level at which severe distortion occurs.
Hertz (abr. Hz) - The unit of measure for frequency. One Hertz equals one cycle per second. (After Heinrich Rudolf Hertz, 1857-1894.)
High Pass Filter - An electronic circuit that allows all frequencies above a fixed frequency (not zero) up to infinite frequency. High pass filters are useful for eliminating low-frequency rumble.
Hi-Z - Symbolic expression for high impedance. In the mathematics of electronics, “Z” is used to denote impedance. When you see the phrase hi-Z input, it mainly refers to a direct instrument input (guitar, bass etc). (See DI).
Impedance - A combination of factors that restrict current flow in an AC electrical circuit. Similar in concept to resistance, impedance has two components: resistance and reactance. Reactance also has two components, a “real” part (resistive), and an “imaginary” part, which is resistance induced by phase shift (e.g. the time it takes for other components such as inductors and capacitors to charge and discharge).
Intermodulation - Describes distortion that occurs as a product of nonlinearities in an audio system that produce beat frequencies in complex waveforms. A beat frequency (the pulsing effect you hear when two notes of the same pitch are slightly out of tune) is the product of frequencies that are not harmonically related to the fundamentals. Rather than being multiples of the fundamentals, they are sum and difference frequencies. Intermodulation occurs when the frequency sum or difference of a particular signal interferes with that of another signal. Overall, intermodulation is a very unpleasant, non-musical form of distortion.
kHz - Hertz in multiples of one thousand are expressed as kHz. (Lower case “k” is the accepted symbol for 1,000.) For example 2.5kHz would equal 2,500 Hertz.
Latency - The time it takes for a signal to pass through a device or to be processed by software. In DAW recording, latency becomes an issue especially during tracking due to a significant delay between the time a note is struck and then reaches the performer’s ear. The combination of hardware and software latency in a computer-based DAW can render performance impossible, since musicians rely on instantaneous feedback for timing and feel. (See Zero-Latency)
LED - An acronym for “light-emitting diode.” An LED is a solid-state device that emits incoherent light while conducting current in a forward direction. Diodes fall into the category of semiconductor, since they only pass current in one direction. Invented by Nick Holonyak Jr. in 1962, LEDs are used in a number of applications including audio metering and on/off indicators. For example, the cool blue, green, yellow, or red lights that come on when you power up your monitors are LEDs.
Limiter - Basically like a compressor, except that it does not allow the signal to go above the setting of the threshold. For example, if the limiter’s threshold (level at which the limiter will act on the signal) is set to –20dB, once the signal level reaches –20dB and beyond, the output level remains at –20dB. Limiters are used to prevent signals from clipping or overloading speakers, recorders, and power amplifiers, etc.
Line Level - The standard operating voltage level for professional and consumer audio equipment. Professional equipment operates at +4dBu, which is 1.23 volts rms, while consumer and some professional equipment operates at –10dBV, or 0.316 volts rms. A –10dBV system will provide equal quality to a +4dBu system. Troubles occur when you try to connect a +4dBu output to a –10dBV input and vice versa. As long as all of the linked units are operating at the same level, joy and merriment will abound.
Linear - An audio system is considered linear whereby the change in output is directly proportional to the input and can be represented on a graph as a straight line. A linear system must demonstrate proportionality, additivity, and behave in a predictable fashion. To test proportionality, a signal is input into the system. If the strength of the signal is doubled, then we can predict that the output should also be doubled. For example, if we mic a guitar amp and the guitar becomes twice as loud, the microphone should respond twice as much if it is a linear system. To test the additive function, a second guitar amp is placed in front of the mic. The response of the mic should equal the sum of both guitar cabinets (M out = G1 + G2). The predictable behavior of linear systems is of particular importance with regard to speakers. When music is played back on a system that exhibits greater variance than the speakers used to mix the music (and vice versa), the levels of certain instruments will appear to change.
Loading (Load) - In electronics, a load consumes energy to do some form of work. Any component or device that is connected to a source and draws current from it is considered a load. Loading is an effect that is a function of the load’s impedance. For example, if a load has high impedance (resistant to current flow) it will draw a small amount of current from the source, thus the load is small (light loading). Conversely, a small amount of load impedance will draw higher load current from the source (heavy loading).
Low-Pass Filter - As opposed to a high-pass filter, a low-pass filter allows all frequencies below a specified frequency to pass while attenuating all frequencies above.
MIDI (Musical Instrument Digital Interface) - Driven by Dave Smith in the early ’80s (founder of Sequential Circuits and designer of the legendary Prophet 5 synthesizer among others) after meetings with Tom Oberheim of Oberheim Electronics and Ikutaro Kakehashi of Roland, the purpose of MIDI was to establish a common communications protocol between electronic instruments and peripherals from different manufacturers around the world. Today, MIDI is used in all aspects of digital music including composition, performance, sequencing, sound and light triggering, synchronization, and more.
Midrange - In describing the frequencies of the audio spectrum for practical purposes, such as in applying equalization, it has become common to divide it into three sections: low, mid (midrange), and high. Since the midrange encompasses a wider range of frequencies than the low or high end of the spectrum, it is also divided into three sections with lower mid and upper mid as qualifiers. (We don’t say, “middle mid.” You can if you want to, but we generally don’t.) The lower mids extend from around 250Hz to 1kHZ and the upper mids range from around 4kHz to 8kHz. Keep in mind that these are just approximations and that there is always overlap. As an aside, even though the upper midrange seems to go pretty high, it just makes more sense to say, “upper mid” than “lower high.”
Mixdown - This is the act of summing or mixing multiple mono tracks in a recording and combining them into one, two, or more (in the case of surround) tracks. And what do we use to perform mixdown? Hands . . . anyone . . . Bueller? Bueller? (Hint: look directly below.)
Mixer - As the name implies, a mixer takes a number of input signals on individual channels and sums (combines) them into one or two output signals. (See Recording Console)
Modeling (Physical Modeling Synthesis) - The use of computational processes including mathematical equations and algorithms to emulate a physical sound source such as a violin, piano, or synthesizer. For example, a virtual instrument software plug-in can accurately produce the sounds of a classic synthesizer, such as a Mini Moog, by modeling not only the output sound, but the effects of the hardware circuits on the sound, such as the tone generator, envelope generator, and even the action of individual components such as the effect of a light bulb placed in the circuit path. As such, modeling is known for its accuracy in terms of reproducing the quality of any sound-producing or sound processing device. While modeling is not a new concept, it has only come to the forefront of music technology in the last two decades thanks to the refinement of the Karplus-Strong Algorithm by Julius O. Smith III (and others), and the increase in available DSP power in the late ’80s.
Mono - Short for monophonic or monaural. A system that takes one or more sound sources and either records, reproduces, or transmits them via a single channel.
Nonlinear - A nonlinear system is one whose behavior is not the sum of its parts or their multiples, and as such, its behavior cannot be accurately predicted. For example, in a nonlinear audio system, the input would not match the output, so that if a pure sine wave were transmitted through the device, its shape would be different at the output.
Nyquist - Theory While working at Bell Labs in the 1920s, Harry Nyquist discovered the criteria for sampled data systems as we’ve come to know them today. Nyquist theory states that for periodic functions (recurring at intervals of time), if you sample at a rate that is at least twice as fast as the signal of interest, then no information (data) is lost upon reconstruction. Since audio waves were proven to be periodic functions by French mathematician Joseph Fourier, they can be sampled without loss of information by following Nyquist’s rules. According to Nyquist’s theory, the highest frequency that can be accurately sampled is one-half of the sampling frequency (known as Nyquist frequency). For example, the Nyquist frequency of an audio CD (Compact Disk) is 22.05kHz, which equals one half of the standardized sampling frequency of 44.1kHz.
Odd-Order Harmonics - Are frequencies that are multiples of a fundamental frequency by odd numbers, such as 3, 5, 7, etc.
Op-Amp (Operational Amplifier) - The basic building block of analog signal processing, an op-amp is a solid-state, integrated circuit that has two inputs of opposite polarity and one output.
Perfect Octave - In western music, an octave is an interval that spans eight diatonic scale degrees above or below a given note. The upper note of a perfect octave has a ratio of 2:1; or in other words, a frequency that is twice that of the lower note. The octave is the first harmonic of the overtone series.
Phantom Power - In conventional DC-polarized condenser mics, phantom power supplies the voltage required to polarize the mic’s transducer element (capsule) using the same two lines as the balanced audio path. It is called "phantom" powering because the supply voltage is effectively invisible to any balanced microphone. The Neumann microphone company developed phantom power in 1966. (See sidebar; Origins of Phantom Power)
Phase - A particular value of time for any periodic or cyclic function, that refers to what fraction of the cycle the signal has advanced to.
Phase Shift - Generally when we speak about phase shift in musical terms, we are referring to a displacement in time between two identical waves. The relationship between the two waves is expressed as an angle taken from a particular point in each wave’s cycle as measured against a specified reference point.
Plug-In - A plug-in is basically a “smaller” program that runs within a host program that provides additional functionality or capabilities to the host software. In audio, plug-ins are generally software programs that emulate the functions or sounds of hardware counterparts, such as compressors, equalizers, reverb/delay units, and even instruments such as synthesizers, grand pianos, and strings etc. While most DAW recording software comes with plug-ins designed by the software’s manufacturer, users can choose to use compatible plug-ins designed by third-party manufacturers instead. Since plug-ins do use a considerable amount of RAM depending on the type of processing, some manufacturers offer “powered” plug-ins, that include software and a dedicated processor card that either plugs into, or connects to a computer.
Polarity - What we call positive and negative electrical potential with respect to a referenced potential. Polarity is often confused with phase reversal. Phase is a function of time, whereas polarity is not. Often in audio engineering we say a signal is out of phase, when in reality we mean its polarity is inverted.
Preamplifier - The first amplifier in the chain, the preamplifier takes a low level signal from a guitar pickup, mic, or turntable, etc., and amplifies it. Technically speaking, the preamp provides significant voltage gain and small current gain, which makes a preamp good for recording applications. A power amplifier must follow in order for current to be amplified enough to power loudspeakers.
RCA Connector (Phono Plug) - Originally developed by the RCA Corporation, an unbalanced pin connector that is now the standard connector used in line-level consumer and project studio sound equipment. (See Cable Buying Guide)
Re-Amping - A recording technique in which an instrument such as a keyboard that was recorded direct (see Direct Recording) is sent to a miked instrument amplifier. The signal from the amp is recorded again and subtly combined with the direct-recorded track in mixdown. The result is the illusion of live performance, while room ambience, amp, and mic response add dimension to an otherwise monochrome sound. Sometimes mic preamps are also used for re-amping. For example, a vocal track recorded through a transparent solid state preamp may be re-amped through a tube preamp to add more body to the voice and subtle compression to the high frequencies.
Recording Console - Takes the basic concept of a mixer and adds a greater number of inputs, signal processing, and routing to multiple outputs in order to feed the individual channels of a multitrack recorder (as well as stereo outputs to feed a 2-track master recorder). Recording consoles also have auxiliary inputs and outputs (aux send/return) and insert jacks that allow signals to be processed by external equipment and returned back to the console before the output stage. Modern large-format recording consoles, (also called desks) will feature a master section that can route summed signals to multiple sets of monitors and control the levels of aux sends and returns, as well as talkback systems that allow the engineer to communicate with artists.
RMS (Root Mean Square) - The square root of the mean (average) of the squares of a group of numbers; rms provides a meaningful way to measure and express the average of a number of discrete values or a continuously varying function. In audio electronics, we use rms to calculate the average of varying voltages, watts, etc. into one useful, meaningful number. For example, in power amplifiers, as different input voltages pass through the system, rms allows us to express the output power in one number that is the average of the varying input levels (e.g. 250 watts rms). Rms is especially useful when variants are positive and negative as in the example of audio waves.
Saturation - The state when a material cannot absorb a stronger magnetic field, such that by introducing a greater magnetization force, no significant change on a magnetic field’s ability to exert a force on a moving charge occurs. In recording, tape saturation is the point at which magnetic tape has reached its maximum ability to accept signals, beyond which distortion occurs. As much as tape saturation is a thing to be avoided, it can be used as a creative tool. When program material is recorded at levels somewhat above 0dB, distortion is being added gradually, which has a pleasing effect. The upper frequencies become gradually compressed as tape is saturated, which also is more natural and pleasing to our ears.
Shelving EQ - In shelving EQ, all frequencies above or below a specific frequency, are boosted or cut equally across their spectrum forming a flat "shelf." Shelving EQ is applied to either the high-frequency range or low frequency range. Bass and treble tone controls are examples of shelving EQs. These controls are good for making broad tonal adjustments, such as adding or reducing bass or treble, but not for making precise adjustments to specific segments in the audio spectrum.
Sibilance - Refers to the hissing “s,” “sh,” “z,” or “zh” sound of the human voice as picked up by a microphone and later amplified. Sibilance is most commonly associated with but not limited to the upper frequencies, which is why many de-essers (devices that manage sibilance) have a selectable frequency range from 800Hz to 8kHz.
S/MUX (Sample Multiplexing) - Proprietary technology licensed by Sonorus, Inc. used to transmit high-bandwidth digital audio using existing lower bandwidth technology. It works by splitting samples of high-resolution signals to multiple channels. For example, an ADAT optical interface is capable of transmitting eight channels of digital audio at a maximum of 24-bit/48kHz. By using the S/MUX protocol, a high-resolution 96kHz signal can be split into two channels of 48kHz audio (S/MUX2). These two channels are combined to create one virtual channel of 24-bit/96kHz audio. Using the S/MUX2 protocol, an eight-channel ADAT optical interface can transmit a maximum of four channels of 24-bit/88.2kHz, 96kHz audio. (S/MUX4 can transmit two channels of 24-bit/176.4kHz, 192kHz.)
Spatial Information - Describes the transient and high frequency components of sound that allows our ears to determine where sounds are coming from (localization). It is also used to describe how accurately or distinctly speakers place recorded instruments in the stereo sound field.
S/PDIF - Sony/Philips Digital Interface. S/PDIF is the consumer version of the AES/EBU serial interface and is also used to transfer digital information. S/PDIF operates at a lower voltage and uses 75 Ohm coaxial cables (up to 30 ft.) with RCA-type connectors and jacks or optical fiber terminated with a TOSLINK (Toshiba link) connector. S/PDIF interfaces can also be found on “prosumer” audio equipment such as the Digidesign DIGI 003 audio interface.
SPL - An acronym for Sound Pressure Level. Measured in dB, SPL is a measure of intensity.
TDIF (Teac Digital Interface Format) - TASCAM’s eight-channel digital audio interface for their DA-88 digital multitrack recorder, which uses an unbalanced 25-pin connector.
Timbre - The unique or subjective quality of a sound that distinguishes it from other sounds regardless of pitch or volume. Timbre describes the quality of sound that allows us to differentiate a flute from clarinet, or one clarinet from another. The timbral quality of a sound comprises a number of factors including attack, harmonic content, sustain, and physical materials, etc. Timbre can be changed by amplifying or attenuating frequency, volume, overtones, and harmonics, even by changing physical location in a room, or via other types of signal processing.
TOSLINK - An optical interface for digital audio connections from Toshiba that provides a data rate from 125 Mbps to 1.2 Gbps. A TOSLINK port is widely used as the optical option of the S/PDIF serial interface found on CD/DVD players, TVs, and other audio and video equipment.
Transformer - A transformer is a passive component comprising two or more magnetically coupled coils or windings (usually wound around magnets) that use electromagnetic induction to increase or decrease voltage and current. Each coil or inductor shares a magnetic path. The more closely they are coupled magnetically, the more efficient they become. (See Electromagnetic induction.)
Transformer-Coupled (also referred to as inductive coupling) - Describes employing a transformer to bridge, or coupling two circuits by means of electromagnetic induction. The advantage of transformer coupling is electronic isolation, which results in a reduction of noise caused by 60-cycle (AC) hum and other sources. Isolation occurs because the transformer eliminates a hard-wire connection between circuits, thus eliminating the possibility of ground loops.
Transient - The leading edge of a waveform at the peak level of the attack portion of a sound (see Attack). As the word implies, a transient exists for a very short duration and is the point at which the loudest part of the sound transitions to the lower volume of the sustain portion of the sound. (See Envelope)
Wave/Waveform - A vibration of atmospheric pressure that moves outward from a source (imagine a stone dropped into a pool of water) and dissipates energy as it travels. When the wave strikes a surface (such as the sides of the pool), it causes reflections, which interfere with subsequent waves. With reference to a neutral point or relaxed state, a wave will have a positive excursion (or compression of the atmosphere in reality space) above the neutral point and a negative excursion (or rarefaction of the atmosphere) below the neutral point. When we use the moving diaphragm of a microphone to convert the wave to electrical energy, or an audio signal, we send that signal through an oscilloscope, which compresses the wave so that we can see it as well as hear it. What we see on the scope is the waveform. Taking a simple wave such as a sine wave, the wave’s positive excursion looks like the top half of an egg, while the negative excursion will look like the bottom half of the egg. Its excursion starts when the downward slope of the positive excursion reaches the median line or neutral point.
Wavelength - The physical space occupied by one complete cycle of a wave. For example, sound travels at approximately 1,100 feet per second (with variations based on altitude and temperature). A 1kHz (1,000 cycles per second) sound has a wavelength in physical reality of a little over a foot.
Weighting Filters - These are a special type of band-limiting filters used to measure varying levels of loudness. Designed to mimic the response of the human ear, they are mostly used to measure noise in audio equipment. The filter emphasizes or “weights” certain frequencies in order to obtain measurements that correlate with the subjective perception of noise.
Word Clock - A synchronizing signal that indicates the rate at which sample words (sampling frequency) are transmitted over a digital audio interface. When interfacing different types of digital audio equipment, such as an audio interface with a computer based sampler, synchronizing via word clock prevents clicks and pops from appearing in the audio. This occurs because the digital converters of each system operate under their own clocking signal, which will have minute differences in sample rate. Slaving the devices to a master word clock prevents this.
XLR - The original model number for 3-pin circular connectors invented by Canon, which has now become a generic term. XLR connectors are used to make balanced cables, such as those used for connecting microphones to preamps. (See Cable Buying Guide)
Zero-Latency - A DAW monitoring scheme designed to circumvent latency introduced by digital signal processing through software and hardware. To achieve zero-latency monitoring (actually near zero) an analog signal input to an audio interface is sent simultaneously to an analog output and to the A/D converters of the audio interface. The analog output may be connected directly to a headphone amp or mixer, so that the performer hears with no delay. Most audio interfaces also have dedicated headphone outputs for this reason.