How Do I Get Rid Of The Noise In The Multitracks Of My Mix?

Noisy multitracks often lead to poor mix results, either from the unhindered passing of that noise into the mix or through the inappropriate processing we apply in trying to minimize the noise. And while noise often can't be eliminated entirely, there are strategies we can employ to reduce noise and achieve clearer, more professional-sounding mixes.

In this article, I'll go into more detail on audio noise, the various types, and how to handle each of the specific types of noise. I'll also cover a few bonus issues we may want to address when it comes to unwanted information in our multitracks, along with a brief section on why noise may actually benefit our mix after all.

---

What Is Noise, And What Types Of Noise Are There?

Noise, in audio signals, is loosely defined as any information that is not directly related to the intended contents of the signal or, in other words, any unwanted information within the audio signal.

Of course, this definition is highly subjective. One man's noisy trash is another man's shiny treasure.

So, I've found it more useful to pin down certain types of noise that can be more easily defined rather than relying on the broad description mentioned above.

The different types of noise (generally speaking) include, but are not limited to:

• Mechanical noise
• Ambient noise
• Electromagnetic interference noise
• Radio frequency noise
• Internal circuitry noise
• Bleed
• General artifacts

Additionally, noise can be continuous or intermittent, and we'll generally deal with these types of noises differently.

Let's briefly run through the general types of noise here.

What Is Mechanical Noise?

Mechanical noise is the noise that is transferred through solid material and into the microphone (or another transducer, such as an electric pickup). These vibrations often trend toward the low end of the audible frequency spectrum.

Examples of mechanical noise could be footsteps around the mic stand, touching the mic stand, mic or pop filter, or anything that would cause strong vibrations to transfer to the transducer and be converted into audio. The examples extend to the low-end rumblings that come from HVAC systems, traffic and construction, vibrating through studio walls and floors and ultimately into our audio signals.

What Is Ambient Noise?

Ambience noise is the noise that travels primarily through the air that is picked up by the microphone or other transducer and converted to audio. These sounds would be captured whether the intended source produced any sound or not.

Examples of ambient noise include HVAC systems or any fans that may be running (including the fans of our computers), along with noise from our appliances and everything happening outside the studio that happens to find its way into the studio to get picked up.

Ambience also includes the natural reverberation of the acoustic space that will surely get captured by a microphone. In some instances, rooms are chosen specifically for their sonic character. In other instances, rooms are heavily treated to sound as sonically “dead” as possible”. In other instances, we'll have less-than-ideal situations where the room yields unwanted resonances, flutter echo, comb-filtering and other poor acoustic results.

Outside the studio, ambience also included wind noise and any other noises that happen outdoors.

What Is Electromagnetic Interference Noise?

Electromagnetic interference (EMI) noise is noise introduced into an audio circuit from an outside source. The devices and cables within the audio path create an electrical circuit and act, in a way, as antennae for electromagnetic interference.

The most common example of EMI noise in audio comes from the AC power feeding our equipment. Depending on where you are in the world, this could mean having the dreaded 60-cycle (60 Hz) hum or the dreaded 50-cycle hum. In many cases, the harmonics of these fundamentals will also find themselves in the signal. That said, any strong enough electrical sources can cause EMI in our audio.

What Is Radio Frequency Noise?

Radio frequency (RF) noise is a specialized form of EMI that picks up radio frequencies (ranging roughly from 20 kHz to 300 GHz). These electromagnetic waves are often used as carriers for audio signals (AM and FM radio), and the modulating audio signals can sometimes be picked up in our audio lines within the studio.

Radio frequency noise is particularly an issue when we're using transducers in proximity to radio stations. For example, a studio I worked at was directly across the street from a TV and radio station, and sometimes the Neumann U87Ai microphone would act as an antenna for the radio signals. As another example, I used to live across the water from Quebec, and my electric guitar pickups would regularly pick up a French radio station, even though it was roughly 15 kilometres away.

What Is Internal Circuitry Noise?

Internal circuitry noise, often referred to as self-noise, is the noise produced by the electrical components of an audio device.

Beyond any obvious parts of a given audio device design (fans, switches, etc.) which produce ambient/mechanical noise, we do have to take into account the noise produced by the electrical components of analog audio equipment.

Passive components like transformers, inductors, capacitors and even resistors can introduce noise into the audio signal as electricity flows through them. Furthermore, active components (ICs, transistors, vacuum tubes, etc.) and their power sources can also introduce noise.

Without getting into the details of electromagnetic induction, Johnson-Nyquist noise, and everything else that goes into the noise produced by electrical components, we can basically state that so long as electrons are moving, there will be noise.

What Is Bleed?

Bleed is the sound of any sound sources other than the intended source, particularly when recording with a microphone. When multiple musicians are individually spot-miked (or even group-miked) and recorded simultaneously in the same room, each microphone will surely capture some sound from the other musicians and sound sources that aren't from the intended source(s).

Any sound source that the microphone is not primarily set up to capture is considered bleed.

For example, a tom drum channel, which is generally recorded with a close-miked setup, will have a lot of bleed from the other drums in the acoustic drum kit, along with any other instruments with sound waves reaching the tom mic.

As another example, even if we have a vocalist in an isolated environment, the vocal microphone may pick up sound from the headphones used for monitoring during the vocal performance. This is “headphone bleed”.

What Are Artifacts?

Artifacts are loosely defined as any undesired audio information that is introduced into the audio signal(s) throughout the music production process (during recording, routing, editing, processing, mixing, mastering, bouncing/exporting, etc.).

These noises are often caused by technical issues and can be described as clicks, crackles, pops, clipping, plosives, etc.

---

The Noise Floor And Signal-To-Noise Ratio

Before moving on, let's quickly discuss the noise flow and signal-to-noise ratio.

The noise floor is the base-level noise of a recording produced by the sum of all the noise sources present in the signal.

The signal-to-noise ratio of a signal is the ratio between the intended signal (the primary sound source being miked, the electric instrument connected to the electric pickup, the synthesizer producing audio, etc.) and the noise of the audio signal.

Fortunately, noise isn't a major issue in digital systems. For example, modern plugins don't typically add nearly as much noise as their analog counterparts, and even the more affordable ADC and DAC (analog-to-digital converters and digital-to-analog converters) don't produce significant noise.

That stated, we still have to concern ourselves with the analog portion of the multitrack-creating process, even if we're working primarily in digital.

One of the big issues with noise is that it will get amplified in every gain stage along with the intended information of the signal.

And so, it's paramount that we strive for low noise when recording our multitracks to get great results at the source.

---

How To Reduce Noise During Recording

The best way to get rid of noise in the multitracks of your mix is to ensure there's no noise, to begin with. Of course, if you're only working as the mixing engineer, there's not much you can do other than request a re-record (which may not go your way). However, if you're the recording engineer, here are a few tips to help eliminate (or at least reduce) the noise during recording.

• Solve acoustic issues
• Solve ambient noise issues
• Solve mechanical noise issues
• Solve bleed issues
• Solve grounding issues
• Solve equipment issues
• Solve RF issues

Solve Acoustic Issues

Acoustic issues may not necessarily be the most obvious type of noise, but I'd like to touch on them here. Even though I'm only scratching the surface, this is going to be a fairly long section, so please feel free to skip ahead to the sections on reducing noise in the multitracks if you'd like!

Acoustics are, by nature, inexorable. Even in anechoic chambers, we have acoustics (an ideal free-field environment).

Now, we generally won't be using anechoic chambers for music production, even if we could afford them in our home studios. So then, the issues of acoustics are perhaps the most difficult to eliminate. However, we can certainly do things to reduce their impact on our multitracks and our final mix.

First, let's consider our acoustic space and our positioning within the space.

Ideally, we want the least-square-shaped room possible. Assuming two pairs of parallel walls (not an uncommon feature in building design), we want different dimensions of height (floor to ceiling), width (parallel pair of walls 1) and depth (parallel pair of walls 2). Cathedral and vaulted-style ceilings are great, as are non-parallel walls, but we may not have spaces with those acoustic “luxuries”.

Note that we also don't want a room where one or two of these dimensions are interval multiples of another.

The reason we want these unrelated dimensional differences is primarily to reduce resonances and anti-resonances caused by standing waves. To keep things simple (acoustics are deeply complex), standing waves are caused primarily by axial resonances, those between two parallel surfaces. The wavelengths that fit between the parallel boundaries (those that are fractions of the distance between the parallel surfaces) will become overly enhanced at certain spots (due to the antinodes) and completely eliminated at others (due to the nodes).

Consider the following diagram with nodes in pink and antinode in red (maximum pressure) and green (maximum rarefaction):

So, standing waves will produce resonances and anti-resonances, and we have to consider where the hot spots and dead spots will be when recording. If we happen to place a microphone directly in the centre of the room or very close to a corner, we can expect a significant build-up of resonant frequencies in the audio signal.

Second, let's consider what's in the room and acoustic treatment. There are plenty of options that we can opt for to improve the acoustics of our recording environment to reduce flutter echo, unruly resonances, excess reverb, comb filtering, and the hot spots (and dead spots) of bass frequencies within the room.

When it comes to acoustic treatment, we're most concerned with two things: absorption and diffusion.

Absorption is when the energy of the longitudinal sound wave (which causes the vibration of air molecules it passes through) is absorbed (primarily due to friction and thereby converted to heat energy).

The air naturally absorbs energy from the sound waves. That's a primary reason why more distant sound sources are quieter. However, in studio environments, we need to account for the surfaces of the acoustic space.

A sound wave emanating from a sound source will travel through the acoustic space (low frequencies will be less directional) until it hits a boundary. If the boundary is flat (more on this in a moment), some of the sound wave's energy will be reflected at the angle of reflection (an angle equal to that at which it approached the surface against the perpendicular of the flat surface).

The amount of energy that is reflected back is defined, in part, by the absorption characteristics of the surface and also by the amount of sound transmission allowed by the surface and the material beyond the surface (sound transmission is the energy that passes through a boundary (in most cases, a wall, floor or ceiling).

Increasing the absorption of a boundary will help reduce the reflections in the acoustic space, thereby reducing issues of flutter echo, resonances, excess reverb, comb filtering, and hot and dead spots.

This is where acoustic panels and bass traps come into play. We can opt to purchase panels ourselves. I personally use the Primacoustic London 12 system. We can also choose to build these panels ourselves.

Simply adding more material to the room will help as well, particularly if it's soft and absorptive. Couches, curtains, rugs, blankets, pillows, and other household furniture can help to reduce acoustic issues with absorption.

In addition to acoustic absorption, we also have diffusion, which acts by breaking up the reflections of a sound wave rather than attempting primarily to absorb the energy. Due to the stepped surfaces, what would otherwise be a simple reflection on a flat surface becomes many diffuse reflections with much less energy.

We can purchase diffusers or make our own. I've found that bookshelves work tremendously well, so long as they're full of books of varying sizes.

Alright, I'm done with my spiel on acoustic noise for this article.

Solve Ambient Noise Issues

Ambient noise issues are perhaps the most difficult to reduce because we often don't have direct control over them.

We can't control the traffic or construction outside, the other tenants in our building (if we're sharing), the other people within the studio, the weather (notably wind and raid) and wildlife (notably birds).

We could, in some cases, turn off our HVAC systems periodically during recording. We can even go as far as turning off our appliances if need be.

There will be some trade-offs when it comes to ambient noise. Our best bet is to acoustically isolate our recording space, though that's much easier said than done. Remember that I mentioned sound transmission in the previous point on acoustic noise — it can get quite burdensome to mitigate sound transmission into a recording space. De-coupling with “floating rooms” and thickening walls, floors, and ceilings can get expensive quickly.

So do your best to isolate your recording environment from ambient noise, but know that, in most cases, it's unavoidable.

Solve Mechanical Noise Issues

Solving mechanical issues is easy in some cases. Avoid hitting the transducers and around the transducers, and you'll stop a lot of mechanical noise.

Try to de-couple the transducers from mechanical vibration. For example, shock mounts are commonplace for holding microphones, and we naturally de-couple electric guitars and basses when we hold them.

The mechanical noise that is more difficult to avoid is that which is ambient—the low rumble of traffic or weather or anything else that we don't have full control over.

Solve Bleed Issues

The simplest way to solve bleed issues is to record things separately.

This could mean recording each instrument separately in succession. For example, we could record all the drums at once to a click track, then overdub the bass, then overdub the guitars and other instruments, and finally overdub the vocals.

It can also mean recording things at the same time but having the sound sources in different, isolated locations. If we have the space and routing capabilities, this can be a good option. For example, we can have most of the musicians all in the same room with their own headphone mixes so they can see each other but have the vocalist in their own iso-booth and the miked-up instrument amp cabinets in their own isolated rooms.

However, in some cases, we want a true live-off-the-floor feel where all of the musicians are in the same room, which makes bleed inevitable. In this case, we can opt to position the sound sources, whether acoustic or amplified, as far as possible. We can get as many DI signals as possible. We can also utilize gobos and other acoustic barriers within the room to help mitigate bleed without obstructing the view and communication between the musicians.

Alternatively, we could produce as much as possible without microphones “inside the box” with virtual instruments and samples.

For example, we could record electric instruments directly into our interface and run them through amp simulations. We could program our drums rather than setting up and recording a real acoustic kit.

This alternative method may work for some, but in many cases, it's not what the music production in question is about. I just wanted to offer it as an option for those of you who could make it work.

Solve Grounding Issues

Grounding issues can lead to hum and other EMI noise in the audio signal. Most often, poor grounding will cause the dreaded 50/60-cycle hum.

With that important disclaimer out of the way, we can reduce the potential for ground issues by connecting everything to a common ground. That means, if possible, connecting the entire studio to a single breaker/fuse's worth of outlets.

Using a power conditioner is another excellent and safer way to do this.

Additionally, if a particular piece of gear is giving you EMI issues and it has a ground lift switch, consider flipping it and listening to hear if the problem solves itself.

Solve Equipment Issues

Sometimes it's the equipment that is inherently noisy. Faulty or dirty connections can lead to crackling, popping and other artifacts, so I recommend starting with cables first if you're having these issues.

If a piece of gear is sounding distorted, it could be that it's being fed with too much level, so bringing the levels down before its input can solve the issue — as an aside, don't feed line level signals into mic level inputs. Otherwise, the internal amplification stage(s) or power supply could be damaged or simply worn out.

If there's significant hum or hiss, and grounding issues aren't to blame, it could be that the gear is damaged and may require fixing.

And sometimes, gear simply has an inherent noise floor that we have to deal with.

Solve RF Issues

Fortunately, solving radio frequency issues doesn't mean we have to move further away from radio stations.

In many cases, simply swapping the microphone and/or cable of the problematic audio channel will do the trick.

However, if that's not possible, you can also invest in an RF blocker like the Shure A15RF that can be inserted in-line to eliminate EMI from radio frequencies.

---

How To Reduce Low-End Rumble In Mix Multitracks

Alright, it's finally time to get into the strategies for reducing noise once it's been recorded into our multitracks.

Let's start with low-end rumble, which is a common byproduct of mechanical and ambient noise. We can almost lump 50/60-cycle hum in this category since it's concentrated in the low end (even though it's not technically rumble).

Our primary tool for reducing low-end rumble in our mix multitracks is the simple high-pass filter.

A high-pass filter (or low-cut filter) does what its name suggests: cuts out low-frequency content while allowing high-frequency content to pass. If we set the high-pass filter's cutoff frequency and slope adequately high and steep, respectively, then we can eliminate the low-end rumble from the signal.

---

How To Reduce Resonant Frequencies In Mix Multitracks

Resonant frequencies are tangentially related to noise in the fact that they're often produced as a byproduct of the acoustic environment, and they're often unwanted in the audio signal.

The best way to reduce resonant frequencies in our multitracks is by using an EQ cut. A simple bell-type filter cut will often work fine, though harsher resonances may call for a band-stop or “notch filter”. If the resonance(s) happen to be in the very low end, we can opt for a higher-pass filter instead, and in the unlikely event that there's resonant build-up in the top end, we can use a low-pass filter.

If you can hear a nasty resonance in one of your multitracks, I would advise trying the boost-and-sweep technique to find exactly where the resonant frequency is (or frequencies are).

This is a straightforward technique where we boost a narrow band in a parametric EQ and sweep that band across the frequency spectrum, listening closely for the increased level of the resonance we heard to be problematic in the first place. Once we find the problematic frequency, we can cut it out of the signal with a narrow bell-type cut or a notch filter.

Note that we need to be careful when employing this technique, as any narrow boost, by default, will sound resonant. It's critical that we don't use this strategy as a baseline for EQ but that we keep it specifically for finding and removing resonances.

If the resonance is intermittent, we may want to opt for a dynamic EQ cut, which will reduce the resonant frequency band only when the energy within that band surpasses a set threshold. In this case, we can find the problematic frequency by boosting and sweeping the EQ and then setting the dynamic controls so that it reduces the resonance appropriately.

---

How To Reduce EMI (Hum) In Mix Multitracks

EMI hum is often concentrated at 50 or 60 Hz (depending on where you live) because of the power mains feeding your equipment from the electrical infrastructure.

When we're dealing strictly with this type of hum, we can often reduce it significantly with aggressive high-pass filtering above the power mains frequency.

However, in many cases, especially if there's any saturation/distortion in the line, there will be harmonic content above this 50/60 Hz fundamental. In this case, we can do our best to get rid of the low-end hum but still have issues with the harmonics.

Beyond the power mains, our multitracks can contain other electromagnetic interference across the audible frequency spectrum.

In many situations, this EMI hum will be rather consistent, and there are tools for reducing such steady-state noise with minor negative side effects. Consider the following tools for hum reduction/elimination:

• De-Hum from iZotope RX (what I use)
• Waves X-Hum
• Oxford De-Buzzer from Sonnox Restore
• DeHum from Acon Digital Restoration Suite

Even if the hum is intermittent or dynamic in level, the above-mentioned tools can help to reduce it across the frequency spectrum.

---

How To Reduce Hiss In Mix Multitracks

As previously mentioned, hiss is commonly produced by the components of our audio system and is sometimes referred to as self-noise.

The easiest way to reduce hiss in the mix, if applicable, is to use fewer analog components or plugins that emulate analog circuits.

But when it comes to reducing the hiss printed in our multitracks, we'll have to opt for different strategies.

Since hiss is typically concentrated in the high frequencies, we can utilize EQ to reduce or even eliminate hiss. Low-pass filters, high-shelf cuts and even bell-type filter cuts in the top end can help minimize the sound of hiss in a multitrack, though they will certainly have an effect on the signal within the affected range as well.

And so, while EQ is an important hiss-reducing tool, it can sometimes do more harm than good in a multitrack if we need the top-end in the mix.

Fortunately, if the hiss is relatively steady in the multitrack's audio, there are specialty tools for reducing it. Consider the following tools for hiss reduction/elimination:

• Spectral De-Noise from iZotope RX (what I use)
• Waves X-Noise
• Oxford De-Noiser from Sonnox Restore
• DeNoise from Acon Digital Restoration Suite

Note that hiss (and hum) sometimes come from the playback system (notably either the amplifier(s) and/or speaker(s)), so the issues may need to be solved at this point.

---

How To Reduce Bleed In Mix Multitracks

Bleed is inevitable if there are multiple sound sources in the same acoustic space.

Perhaps the most common tool for reducing bleed in a multitrack is the noise gate. This is especially true for sources that aren't consistently producing sound throughout the entirety of the song (think tom drums, vocals, and other intermittent instruments).

A noise gate is a dynamics processor that effectively mutes a signal as the sidechain level drops below a set threshold. With the input signal acting as the default sidechain, a noise gate will “close” (mute) the audio as its level drops and “open” to allow the signal to pass through when there's sufficient signal level (above the set threshold).

By adjusting the threshold and time controls, we can use a gate to allow certain parts of a signal through while muting the parts with low signal-to-noise ratios.

Alternatively, we can gate our multitracks “manually” but cutting them along the timeline and muting the sections without significant signal (but with significant bleed). Be sure to add crossfades in this case to ensure no digital clicks and pops or other artifacts.

If the bleed happens to be outside the bandwidth of the main signal within the multitrack, we can have some luck with EQ. This is particularly useful for low-end bleed from kick drums and bass instruments, which can be reduced with a simple high-pass filter. High-end bleed can be reduced with low-pass filters or high-shelf cuts, so long as the EQ moves don't overly affect the intended signal.

If the bleed is causing resonant frequencies within the multitrack, we can opt for notch filters or even dynamic EQ cuts in some cases.

Furthermore, if the bleed is fairly consistent, there are tools to help reduce wideband bleed. Consider the following tools for bleed reduction/elimination:

• De-Bleed from iZotope RX (what I use)
• Waves Clarity Vx Pro (what I use)

---

How To Reduce Wideband Ambient Noise (HVAC, Appliances, Etc.) In Mix Multitracks

The reduction of wideband ambient noise is especially important to get at the source. Recording in quiet environments is key, though some amount of ambient noise is almost certain to be captured in our multitracks.

Like the bleed mentioned in the previous section, if the ambient noise happens to be concentrated outside the bandwidth of the signal, high-pass, low-pass, and even band-pass filters can help reduce it.

However, there's not much we can do with EQ in truly wideband ambient noise.

That stated, we do have access to a few tools to help with this type of noise. If the ambient noise is consistent, we can deal with it relatively easily. Even if it isn't, there are several tools to help us reduce wideband ambient noise with minimal colouration or artifacts. Consider the following tools for wideband ambient noise reduction/elimination:

• Spectral De-Noise from iZotope RX (what I use)
• Waves X-Noise
• Oxford De-Noiser from Sonnox Restore
• DeNoise from Acon Digital Restoration Suite

---

How To Reduce Excessive Reverb In Mix Multitracks

While recording in sonically pleasing environments can be a tremendous help in mixing a record, having excessive or unwanted reverb and echoing in a multitrack is problematic.

One method of reducing excessive reverb in a multitrack is to employ transient shaping. As the name suggests, a transient shaper will alter the transient information of a waveform. In order to reduce excessive reverb in a multitrack, then, we can opt to enhance the attack of the transients and, more importantly, reduce the sustain or “tail” of the sound.

By shaping the transients appropriately, we can effectively reduce the sound of the reverb/echo in the multitrack. However, we likely won't be able to get rid of it entirely or even significantly without producing some amount of unnatural pumping. Be sure to use your discretion.

Consider the following transient shapers for this purpose:

• Kilohearts Transient Shaper (free)
• Plugin Alliance SPL Transient Designer Plus

Beyond transient shapers, there are dedicated tools for reducing reverb in audio files. Of course, these options also come with their side effect, so be sure to use your ears when dialling them in, if you choose them at all. Consider the following tools for reverb reduction/elimination:

• De-Reverb from iZotope RX (what I use)
• DeVerberate from Acon Digital Restoration Suite

---

Bonus: How To Reduce Plosives In Mix Multitracks

Vocal tracks will often have plosive energy that should be reduced or eliminated to best serve the mix.

Since a lot of the plosive energy that will overload a microphone happens to be in the low-end frequencies, a simple high-pass filter is often enough to reduce plosives to an adequate level.

When opting to reduce plosives with a high-pass filter, be sure not to overdo the filtering for the sake of plosive reduction. We still need the power in the vocal. On that note, it's often easier to reduce plosives on female vocals with this method as they are generally higher in frequency content.

Level automation at the vocal plosive is another useful technique. Simply bringing the volume down just before the plosive and then bringing it right back up can eliminate a lot of the overload. This technique may sound a bit unnatural in solo, but it can often work perfectly in the context of the mix.

Similarly, we can opt to automate the cutoff/corner frequency of a high-pass filter to get similar results at plosive points in the vocal without affecting the level of the midrange and upper frequencies whatsoever.

In addition to these fixes, there are also dedicated tools for dealing with plosives in the mix. Consider the following tools for plosive reduction/elimination:

• De-Plosive from iZotope RX (what I use)
• Oxford De-Noiser from Sonnox Restore
• DeNoise from Acon Digital Restoration Suite

---

Bonus: How To Reduce Sibilance In Mix Multitracks

Dealing with sibilant energy means dealing with a specific range of a vocal, typically between about 5 and 8 kHz.

What we want to do here is reduce the energy in this range, but only when it's too harsh. So then, static EQ can reduce sibilance in the vocal, but at the cost of reduced intelligibility at all other times.

To properly deal with sibilant energy, we need dynamic control. This can be done with a dynamic EQ with a dynamic cut in the 5-8 kHz range. It can also be done with a multiband compressor setup with a band specifically targeted at the 5-8 kHz range.

It can also be done with specialized de-essers (which are largely based on narrow-band compressors), which hone in on the sibilance range and reduce sibilants when they exceed a set threshold. Consider the following de-essers for reducing excessive sibilants in the mix:

• Waves DeEsser (what I use)
• FabFilter Pro-DS

---

Bonus: How To Reduce Other Artifacts In Mix Multitracks

Other artifacts we may encounter in our multitracks and mixes include digital clips, digital clicks/pops, and dropouts, among others, that are both rarer and more difficult to fix.

Of course, the best way to fix artifacts is not to have them in the first place. This means ensuring proper recording levels, crossfading, master clock setup, and other best practices. However, we may still find artifacts in our multitracks that we need to fix.

Sometimes, suppose the problematic artifact is in a region that's repeated throughout the multitrack. In that case, we may be able to cut, copy and paste a portion of an unaffected region over the affected region. This can be a good strategy for keeping things sound natural (without processing).

Other times, we may be able to cut the problematic samples out completely and crossfader around them. This strategy can be employed more with background multitracks rather than the main elements.

Some other times, we can actually get away with leaving the artifact in the mix. If the mix is dense enough and the artifact isn't overly problematic, we may be able to leave it in. Some volume automation may be appropriate to bring it down a bit.

Alternatively, we may have to rely on our audio restoration plugins to solve these issues. Look for de-clip, de-click/pop and other restoration plugins to help reduce or even eliminate the effect of artifacts in the multitracks and the mix.

---

Using Noise In Our Mixes

Now that we've spent all that time discussing how to reduce noise in our mixes, let's consider a few reasons why noise may actually be beneficial.

To start, noise is naturally part of our sonic experience of the world. Unless we find ourselves in a perfect anechoic chamber, we will experience some level of noise. We even have our own “self-noise” that we live with.

So if art is to imitate life, then having some amount of noise in our multitracks and mixes is no big deal. In fact, a mix totally devoid of noise (as is the case with music created largely with in-the-box synths and samples) might sound a bit uncanny to us, and sometimes noise is even added to enhance the sense of realism and space in the mix.

Beyond the typical music production techniques that involve the use of noise (noise sweeps before/after build-ups as transition elements, layering noise bursts with drum samples, using noise as a modulation source, etc.), we can opt to utilize noise to add character and life to the mix.

Some low-level noise (notably coloured noise and ambience) can help to fill in the frequency spectrum of our mix and add to the overall dimensionality without masking any of the important elements within the mix.

---

Related Questions

How do I get more clarity in my mix? Here are 10 tips for greater clarity in your mixes:

1. Get the balance right
2. Keep spatial effects relatively quiet
3. Reduce frequency masking
4. Use panning for separation
5. Clean up the low end
6. Pay special attention to the “muddy” low-mid frequencies
7. Maintain transients (don't over-compress)
8. Add presence with saturation
9. Boost a little bit of brilliance or air
10. Consider sidechain compression

What is frequency masking, and how can I reduce it? Frequency masking happens when multiple sounds compete for certain frequency bands and become ill-defined within those bands. Strategies to combat frequency masking include balancing with faders and pan pots, EQ, sidechain compression and arrangement changes.