Why Sounds Mask Each Other

Episode 5 · How Hearing Works · Published Aug 13, 2026

Episode purpose

This episode follows Episode 4's discussion of auditory scene analysis. The previous episode asked how the brain can organise a mixture into separate perceptual objects. This one asks the opposite question: what happens when two sounds compete so closely that the listener cannot hear one of them clearly? The roadmap places these four articles together, with the episode ending by challenging the idea that a frequency chart alone can tell you what will be audible.

Script

Opening

Here's a familiar mixing problem. You have a vocal, guitars, a bass, drums. Everything sounds fine by itself, the vocal sounds clear, the guitar sounds clear, the bass sounds full, the drums sound punchy. Then you press play on the whole mix. Suddenly the vocal doesn't seem as clear. You turn it up. Now the guitar feels too quiet. You turn the guitar up. Now the vocal disappears again. So you EQ the guitar, cut something, maybe it works, maybe it doesn't. You keep adjusting, and eventually you find yourself staring at a frequency analyser trying to figure out exactly where the problem is.

But here's the problem with that approach: a frequency chart tells you where energy exists. It doesn't tell you exactly what the listener will perceive. Two sounds can occupy the same frequency region without necessarily masking each other severely, and two sounds that look quite different on a spectrum can still interfere perceptually.

The reason is that human hearing doesn't analyse frequency as infinitely thin lines. The auditory system groups nearby frequencies into perceptual regions. Those regions are closely related to something called critical bands, and understanding them gives us a much better way to understand frequency masking. Not as "these frequencies cannot coexist," but as "these sounds are competing for the same perceptual resources." That distinction is extremely important.

1. Frequency Is Not Heard as a Perfect Graph

When you look at a spectrum analyser, you see a very precise representation: energy at one frequency, another peak somewhere else, a harmonic series extending upward, a low-frequency fundamental, several resonances. The graph can make it feel as though the ear works exactly the same way. It doesn't.

The auditory system contains the cochlea, a frequency-organising structure in the inner ear. Different portions of the cochlea respond preferentially to different frequency regions, but the responses aren't infinitely precise, nearby frequencies can stimulate overlapping regions. That overlap is one of the foundations of auditory masking. So instead of imagining the ear as a perfect frequency ruler, imagine it as a system containing overlapping perceptual filters. Those filters aren't software EQ bands, they're part of how hearing works.

2. Critical Bands

The term critical band describes an important range of frequencies within which sounds interact strongly in auditory perception. The exact bandwidth changes with frequency, so don't think of a critical band as a fixed number of Hz wide everywhere, that would be an oversimplification. Instead, think of it as a perceptual region.

If two sounds are sufficiently close within the auditory system's frequency resolution, they can compete. If they move farther apart, they may become easier to distinguish. This helps explain something every mixer experiences: two instruments can have overlapping frequency content without necessarily sounding completely confused. The amount and character of masking depend on more than whether two spectrum plots happen to overlap.

3. The Ear Doesn't Give Every Frequency Its Own Private Lane

Imagine a road with infinitely many lanes, one car per lane, never interfering. That would be a convenient model of frequency. But hearing isn't like that. The auditory system has finite frequency resolution, nearby sounds can interact. So instead of infinitely narrow lanes, imagine broader zones. If too much information enters the same zone at once, the listener can have more difficulty distinguishing the individual components. That's the basic intuition behind masking.

4. Masking Is About Audibility

Let's define the term carefully. Masking occurs when the presence of one sound reduces the audibility or detectability of another sound. The important word is audibility. A masked sound hasn't necessarily disappeared from the physical signal, it's still there, the waveform still contains it, the spectrum still contains its energy. But the listener may perceive it less clearly.

This distinction is fundamental to mixing. You can have a vocal physically present in the mix while its words become harder to understand. You can have a bass note physically present while its pitch or articulation becomes less obvious. You can have a percussion transient present in the waveform while another sound makes it less perceptually prominent. The question isn't "is the signal there," it's "can the listener perceive the information I need them to perceive?"

5. Masking Is Not the Same as Frequency Overlap

This is one of the biggest misconceptions. Suppose a vocal contains energy around the same broad region as a guitar. A spectrum analyser shows overlap. Does that automatically mean the guitar is masking the vocal? No. It means they share some spectral territory. Whether that creates a meaningful perceptual problem depends on things like:

• Relative level
• Spectral structure
• Temporal behaviour
• Bandwidth
• Harmonic content
• Register
• The particular sounds involved
• What information the listener is trying to hear

This is why a rule like "two instruments cannot occupy the same frequency" is simply not useful. Music would barely exist if that were true.

6. Overlap Can Be Useful

Think about a piano and a guitar playing the same chord, their spectra overlap, that's part of why they sound like a musical combination. Think about a kick drum and bass, they can occupy overlapping low-frequency territory, and that overlap can create weight. Think about doubled vocals, their spectral overlap contributes to the impression of a larger vocal sound. Think about an orchestra, there's enormous spectral overlap between instruments, yet we hear an orchestral section as a coherent musical object.

So spectral overlap isn't inherently a problem. It becomes a problem when the overlap prevents the listener from perceiving something that matters.

7. The Question Is Always: What Information Needs to Survive?

This is the FREQ way of approaching masking. Don't begin with "where is the overlap." Begin with "what am I trying to hear?" Suppose the vocal is the most important element. You aren't necessarily trying to make it occupy a unique frequency range, you're trying to preserve its words, pitch, articulation, emotional expression, presence, phrase shape.

If another instrument overlaps with the vocal but doesn't interfere with those perceptual cues, there may be no problem. If it reduces intelligibility or causes the vocal to lose its intended hierarchy, then you have a masking problem worth addressing. That's a much more musical definition.

8. A Simple Experiment

Take a vocal and a guitar. Listen to the vocal alone, then the guitar alone. Now listen to them together. Don't look at the analyser. Ask: what information from the vocal became harder to hear? Maybe the consonants, maybe the pitch, maybe the entire vocal feels farther away, maybe nothing meaningful changed.

Now turn the guitar down slightly. Listen again. Did the vocal become clearer? If yes, you've demonstrated a very simple form of masking management. Notice what we did: we didn't start with "find the conflicting frequency." We started with "what became harder to hear?"

9. Critical Bands Explain Why Small Spectral Changes Can Matter

Imagine two tones close together. If they're close enough to interact strongly within the auditory system, changing their relative level can change how clearly one is perceived. A relatively small adjustment can therefore produce a surprisingly large perceptual change.

This is one reason experienced mixers sometimes make very small EQ changes. The goal isn't necessarily to dramatically reshape the spectrum, it can be enough to change the balance inside a region where two sources are competing. That can make one source perceptually easier to follow.

10. But the Same Frequency Can Mean Different Things

Suppose a guitar, a vocal, a piano and a snare all have energy in a certain region. It's tempting to say "that frequency is already occupied." But frequencies aren't reserved seats, the auditory system is hearing patterns. A sustained guitar note and a short vocal consonant may interact differently from two sustained tones. A narrow tonal component and a broadband transient may interact differently. A harmonically rich source behaves differently from a simple tone.

The temporal envelope matters, the spectral shape matters, the relative levels matter, the musical context matters. This is why frequency-only thinking can lead to unnecessary processing.

11. The Stronger Sound Can Mask the Weaker One

A basic masking relationship is level. When a masking sound becomes stronger, it can make a weaker nearby sound harder to detect. This is intuitive: imagine somebody whispering while a loud guitar amplifier is playing nearby, the whisper hasn't disappeared, it's simply harder to hear.

Now translate that into a mix. A quiet backing vocal can become harder to perceive when surrounded by a dense instrumental arrangement. A subtle percussion part can disappear when the surrounding energy becomes more intense. A bass articulation can become less obvious beneath a strong kick. The mechanism is more complicated than simply "louder wins," but relative level is an important part of the problem.

12. Masking Can Change the Character of a Sound

Here's something mixers sometimes miss. Masking doesn't necessarily make a sound disappear completely, it can make part of its character disappear. You may still hear that there's bass, but its articulation becomes less clear, the pitch may be less defined, the attack may disappear while the sustain remains. Or a vocal may still be audible, but consonants become less distinct.

This matters because musical identity often depends on details rather than simply overall audibility. A sound can be technically present while losing the information that makes it useful.

13. Harmonics Make Masking More Complicated

Real musical sounds are not single frequencies. A note has a fundamental and harmonics, and the harmonic series can extend far beyond the fundamental. So when you think about masking, you can't only look at the fundamental frequency of an instrument. A bass note may have substantial harmonic information much higher than its fundamental, a guitar has many harmonics, a piano has a complex harmonic structure, a vocal contains a fundamental plus harmonics, formant structure and consonant energy.

That means two sounds can interact in many different frequency regions simultaneously. This is another reason "find the fundamental and carve around it" isn't a complete mixing philosophy.

14. The Fundamental Isn't the Whole Sound

Imagine a bass note whose fundamental is very low. You might assume the listener primarily hears that low-frequency component, but the perception of pitch and timbre can also involve higher harmonics. This becomes particularly important when discussing playback systems, a bass sound can remain perceptually recognisable even when some low-frequency components are missing. That phenomenon connects to the missing fundamental, which we'll explore more deeply in the kick-and-bass episode later.

For masking, the immediate lesson is: don't assume removing one low-frequency component removes the perception of the instrument. The auditory system integrates harmonic information.

15. Upward Spread of Masking

Now we come to one of the most important ideas in this episode. Masking isn't necessarily symmetrical around the masker. A strong low-frequency sound can affect the audibility of higher-frequency sounds. This tendency is called the upward spread of masking, a strong lower-frequency sound can mask information above it.

This matters enormously in music. A dense low-frequency source isn't only a low-frequency problem, its perceptual effects can extend upward. That's one reason low-mid and bass buildup can make a mix feel unclear in a broader sense than the spectrum might suggest.

16. Why Low-Frequency Problems Can Make a Mix Feel Cloudy

Imagine several sources generating substantial low-frequency energy: bass, kick, low guitar, piano, synth, room information. Individually they may all sound fine. Together, the lower part of the spectrum becomes dense. The result can be more than "too much bass," the mix can begin to feel less transparent, higher-frequency information may become less perceptually prominent, the vocal may feel less open, the arrangement may feel smaller even though there's more energy.

This is one reason a muddy mix can be difficult to diagnose by simply looking for one obvious frequency peak. The issue can be an accumulation of overlapping information.

17. Why 250 Hz Keeps Appearing in Mixing Conversations

This is where the FREQ article on why 250 Hz matters becomes useful. The point isn't that 250 Hz is universally bad, it isn't. It's an important region because many musical sources contribute energy there. Depending on the instrument, arrangement and recording, that region can contribute body, warmth, fullness, weight, or on the other hand boxiness and congestion.

The same energy can be useful or problematic. If several sources accumulate there, the listener may perceive the combined result as congested. That doesn't mean "cut 250 Hz." It means "listen to what information is competing there."

18. This Is Why a Mix Can Become Muddy Without One Track Being "Muddy"

Take five clean tracks. Each one sounds fine in isolation. Put them together, now the mix sounds muddy. How? Because the problem may exist at the level of interaction, each track is acceptable, but the combination creates excessive perceptual density. This is why soloing every track and fixing each one individually can fail, you're evaluating the wrong condition. The listener hears the combination, so masking needs to be evaluated in context.

19. Solo Can Hide a Masking Problem

This is one of the most important practical lessons. Suppose the bass sounds perfect in solo, the guitar sounds perfect in solo, the vocal sounds perfect in solo, everything is beautiful. Now play them together, the vocal disappears.

If you EQ the vocal while listening soloed, you may make it sound excellent by itself. But that doesn't tell you whether it works in the mix. The question isn't "does this track sound good," it's "does this track remain perceptually useful when everything that competes with it is present?" That's the environment in which masking happens.

20. Masking Is Often Context-Dependent

A sound can be perfectly audible in one section and masked in another. Why? Because the competing information changes. The verse may have only piano and vocal. The chorus adds guitars, backing vocals, additional percussion, synths, bigger drums. The vocal hasn't necessarily changed, but its perceptual environment has. This is why something can be fine in the verse and wrong in the chorus, and why that kind of complaint is worth taking seriously as a diagnostic clue rather than dismissing as inconsistency.

21. Masking Can Be Tonal, But It Can Also Be Temporal

This episode is mainly about frequency masking, but remember Episode 6 is coming. Sounds can also interfere with one another because they occur close together in time, a strong transient can temporarily reduce the audibility of something arriving immediately afterward, that's temporal masking.

The distinction is useful. If two sounds are competing because their spectral energy overlaps, frequency masking may be relevant. If one sound's attack is interfering with what happens immediately after it, temporal masking may be more relevant. And sometimes both occur together. That's exactly why the next episode is a listening experiment rather than another theory-heavy discussion.

22. Frequency Masking Doesn't Mean "Use EQ"

This is another important FREQ principle. If two sounds mask each other, EQ is one possible response, it isn't the definition of masking. You might solve the problem through level, arrangement, register, timing, dynamics, panning, depth, or processing.

The best solution depends on why the masking is happening and what the musical intention is. If the guitar is simply too loud, turning it down may be better than EQ. If two instruments are playing unnecessarily similar material, the problem may exist before mixing. If a transient is causing the conflict, dynamics may be more appropriate. If two sounds are supposed to fuse, perhaps the "masking" is actually desirable.

23. Masking Can Create Cohesion

Let's turn the idea around. Some degree of spectral overlap can make sounds feel connected. Think about a rhythm section, if the kick and bass occupy completely separate worlds, the low end may feel disconnected, some interaction helps them become a coherent foundation. Think about doubled vocals, their overlap contributes to the impression that they belong together. Think about layered synths, different layers may occupy similar spectral territory precisely because you're trying to build one larger sound.

So masking exists on a continuum with fusion and integration. A mixer has to decide how much independence is desirable.

24. When Masking Becomes a Problem

A useful test is "what disappeared," not "how much overlap is there." Suppose the vocal is still clearly understandable, the guitar and vocal blend beautifully, you see overlap on the spectrum. Does anything need fixing? Probably not.

Now suppose the vocal's consonants disappear whenever the guitar enters. That's different, a perceptually important piece of information has been lost, and you have a reason to investigate. This is the difference between spectral overlap and meaningful masking.

25. Listen for the Information That Disappears

Take a vocal with a dense guitar arrangement. Listen to the vocal carefully. Don't ask "is the vocal loud enough." Instead listen for consonants, vowel clarity, pitch movement, breath, phrase endings, emotional changes. Now mute the guitar. What suddenly becomes easier to hear? That difference tells you what the guitar was masking. This is much more informative than looking at two EQ curves and deciding they overlap.

26. The Frequency Chart Is Not the Listener

Frequency analysers are extremely useful. They can show you where energy is concentrated, whether something has unexpectedly high low-frequency content, how harmonics are distributed, how two signals differ spectrally, whether an EQ change actually did what you expected.

But the analyser doesn't know what the listener needs to hear. It doesn't know the vocal's consonants are important. It doesn't know the bass needs to feel connected to the kick. It doesn't know the guitar is intentionally supposed to merge with the piano. It doesn't know a particular overlap is part of the sound's identity. It sees energy. You hear meaning.

27. Two Identical Spectrum Shapes Can Sound Different

Two sounds can have broadly similar spectral distributions and still be perceptually very different. A sustained pad and a rhythmic guitar can occupy similar frequency regions but have different temporal structures. A bass and kick can both contain low frequencies while having very different envelopes. A vocal and synth can share frequency regions but have very different harmonic structures. The spectrum is one dimension. Hearing is multidimensional.

28. The Ear Is Not an EQ Analyzer

This sounds obvious, but it's easy to forget. When you listen to a mix, your auditory system is integrating frequency, timing, intensity, spatial information, harmonic relationships, temporal patterns, expectations, attention, context. The analyser gives you a graph. Your auditory system gives you a perceptual scene. The graph can help you investigate, but it shouldn't replace listening.

29. Don't Cut Because a Graph Looks Busy

A busy spectrum can be perfectly musical, an orchestra looks busy, a full drum kit looks busy, a dense guitar arrangement looks busy, a vocal with strong harmonics looks busy. "Busy" is not a diagnosis. Likewise, a clean-looking spectrum isn't proof of clarity, you can have relatively separated frequency regions and still have poor hierarchy, bad timing relationships, excessive dynamics, unclear spatial organisation, or competing musical roles. The ear decides.

30. A Better Diagnostic Workflow

Here's a practical masking workflow:

• Step 1: Identify what isn't being heard. Don't start with frequency, start with perception.
• Step 2: Identify when it disappears. Is it always masked? Only when another instrument enters? Only at higher density? Only at certain notes?
• Step 3: Identify the likely competing source. Mute things one at a time, listen for what suddenly becomes clearer.
• Step 4: Determine whether the overlap is intentional. Do you want fusion, or independence?
• Step 5: Choose the least destructive solution. Try level first, then consider arrangement, register, timing, dynamics, space or EQ depending on the problem.
• Step 6: Re-evaluate in context. Don't judge the solution only in solo, the listener hears the mix.

31. Use Small Changes

Masking problems often don't require dramatic EQ. A small change can alter the balance between two competing sounds enough to restore intelligibility. This is another reason not to think in terms of "remove the frequency." You don't necessarily need to remove anything, you may simply need to change the relative relationship. A guitar doesn't need to lose its character to allow a vocal to become clearer, the vocal doesn't need to become brighter, the guitar doesn't need to become thinner. Sometimes a very small change is enough.

32. The Listener's Attention Is Finite

This connects Episode 4 to this one. Last episode, we talked about how the auditory system organises complex information. Now we're seeing one of the constraints on that organisation: the listener cannot give unlimited attention to everything. When multiple sources compete for similar perceptual territory, the brain has more difficulty maintaining separate representations.

This doesn't mean the brain literally has a fixed number of attention slots, but it does mean perceptual clarity depends on how information is organised. Mixing is partly the management of that competition.

33. Why Dense Arrangements Can Feel Smaller

This is a fascinating paradox. You add more instruments, the mix becomes technically bigger, but it can feel smaller. Why? Because the added information can reduce the perceptual distinctness of the important elements. The chorus contains more sounds, but the vocal is less clear. The guitars are larger, but none of them feels individually important. The low end is fuller, but the bass line is less defined.

The listener receives more information but may extract less useful information. That's perceptual density. We'll explore that properly in Episode 7.

34. Masking and Hierarchy Are Connected

Suppose the vocal is supposed to be the primary object. Then masking of the vocal matters more than masking between two background textures, that's obvious from a musical perspective, not all sounds have equal perceptual importance.

This means masking is also a hierarchy problem. A background texture can mask another background texture without causing a meaningful problem, but if that same texture masks the lead vocal, the consequences are much larger. So the right question isn't "is anything being masked," something almost always is. The question is "is something important being masked?"

35. The Mix Doesn't Need Perfect Separation

Imagine a mix in which every element is completely distinct, you could identify every track instantly, nothing overlaps, nothing fuses, nothing competes. It would probably sound strange. Music depends on interaction: harmonic blending, rhythmic reinforcement, layering, doubling, shared ambience, tonal relationships. The goal isn't to eliminate perceptual competition. The goal is to control it. That's the difference between clarity and sterility.

36. The "Carve a Hole" Mentality

A common mixing phrase is "carve a hole for the vocal." There's some truth behind it, but taken literally it becomes destructive. You don't need to remove everything around the vocal, you need to preserve the information that makes the vocal perceptually useful. That might mean reducing a competing source, changing its dynamics, changing its position, changing its depth, or a very small tonal adjustment. The goal isn't an empty spectral hole. The goal is a clear perceptual role.

37. Masking Can Be Musical

Let's imagine two distorted guitars, they overlap heavily, you might call that masking. But if the goal is a huge wall of guitar, that overlap may be exactly what creates the size. Now separate them dramatically, you can suddenly hear two guitarists, but the combined instrument may feel smaller. So the "masking" was actually part of the musical design. This is why psychoacoustic concepts must always be interpreted through musical intention.

38. A Listening Experiment: Build a Wall

Take two guitar sounds. Start with them completely separate, listen. Now move them closer in register, listen. Make their rhythms more similar, listen. Move them toward the same spatial area, listen. Make their tones more similar, listen.

At some point they stop feeling like two independent guitars, they become a larger guitar object. Now reverse the process, introduce differences. You're listening to the boundary between independence and fusion. Masking is part of that boundary.

39. A Listening Experiment: Find the Masker

Take a dense mix. Pick one element you feel is disappearing. Mute the other instruments one at a time, don't change anything, you're simply asking: which removal makes the missing element suddenly become clear? That source is a candidate masker.

Now bring it back, lower it slightly. Does the missing element return? If so, you've learned something important. You didn't need a spectrum analyser, you listened to the interaction.

40. A Listening Experiment: Find the Region Without Looking

Listen to two sources together. Slowly change the EQ of one, don't look at the frequency display. Listen for the moment the second source becomes easier to hear, then look at what you changed. This is a useful ear-training exercise because you're teaching yourself to associate a perceptual change with a spectral change. But the order matters: hear first, look second. Otherwise the graph can tell you what you're supposed to hear.

41. Upward Masking Listening Experiment

Take a bass-heavy source, play it alone. Then introduce a higher-frequency melodic element. Now increase the low-frequency energy of the bass. Listen to whether the higher element becomes less perceptually distinct. The important question isn't "did the high-frequency waveform disappear," it didn't. Ask "did the higher musical information become harder to follow?" That's the perceptual phenomenon we're interested in.

42. Don't Memorise Critical Band Numbers

You don't need to memorise tables of critical bandwidths to become a better mixer. The useful knowledge is conceptual: the auditory system has finite frequency resolution, nearby components can interact, the interaction depends on frequency and level, masking is not uniform across the spectrum, lower-frequency maskers can produce upward spread of masking, musical sounds contain many components, and the listener perceives the combined result, not your spectrum chart. That's enough to make better decisions.

43. What This Changes About EQ

Once you understand masking this way, EQ becomes less about "fixing frequencies." It becomes a tool for changing relationships. You might reduce one source slightly so another becomes more intelligible. You might emphasise a region that carries important identity. You might reduce unnecessary density. You might leave an overlap alone because it creates cohesion. You might decide the best solution isn't EQ at all. That's a much more mature way of using an equaliser.

44. The Best EQ Decision Might Be No EQ

Suppose a vocal is masked by a guitar. You discover the guitar is simply too loud, turn it down, problem solved, no EQ. Or perhaps the guitar and vocal are playing exactly the same rhythmic figure, changing the arrangement might solve it more effectively. Or perhaps the vocal is too dynamically inconsistent, compression or riding may make it more consistently audible. Or perhaps the guitar needs to move deeper, again no EQ required. Masking identifies a perceptual problem. It doesn't prescribe the processor.

45. Masking Can Exist at Every Depth

Remember the depth discussion from Episode 4. Two sources can occupy similar frequency territory but exist at different perceptual distances, a foreground vocal can coexist with a deeper guitar because the listener receives multiple cues distinguishing them. Likewise, two sounds can be spectrally different but occupy the same perceptual foreground and compete strongly for attention.

So frequency is only one axis. Depth is another. This is why the FREQ approach treats mixing as multidimensional organisation rather than frequency carving.

46. The Bigger Lesson

Let's step back. Episode 4 asked how the brain organises sounds into objects. Episode 5 asks what makes that organisation difficult. Masking is one answer: when several sounds compete within similar perceptual regions, one can reduce the audibility of another.

Critical bands help explain why frequency overlap can matter. Upward spread of masking helps explain why strong low-frequency energy can have consequences above it. But none of these concepts tells you "cut this frequency." They tell you "pay attention to this interaction." That distinction is the entire point.

47. What I Want You to Remember

If you remember only a few things from this episode, remember these:

• Frequency overlap is not automatically masking. Overlap is normal in music.
• Masking is perceptual. The important question is whether useful information becomes harder to hear.
• Critical bands mean the ear doesn't resolve frequency infinitely finely. Nearby spectral components can interact.
• Masking depends on context. Relative level, spectral content, timing and musical role all matter.
• Low-frequency energy can produce upward spread of masking. A low-end problem can have consequences beyond the low end itself.
• Spectrum analysers are useful diagnostic tools, not substitutes for hearing. A chart shows energy, it doesn't show meaning.
• Masking isn't automatically bad. Overlap can create fusion, weight, cohesion and size.

And finally: don't ask which frequencies need to be removed. Ask which information the listener is failing to hear, and why.

Closing

The next time you look at a frequency analyser and see two instruments occupying the same region, don't panic. Music is supposed to overlap. Instead, close your eyes, listen. Ask what you're actually hearing. What disappeared? When did it disappear? Which other sound appeared at the same time? Is that overlap helping the sounds become one object, or is it preventing the listener from following two things that are supposed to remain independent?

Those questions will take you much further than drawing empty spaces into a spectrum. Because the ear doesn't hear frequency charts, the ear hears relationships. And the job of the mixer isn't to make the spectrum look tidy. It's to make the musical information survive.

In the next episode, we're going to move from where sounds overlap to when sounds overlap. Because a loud transient can interfere with something that happens just before or after it. A snare can change how the following sound is perceived. A kick can affect the bass immediately after its attack. And compression can sometimes improve separation not because it changes the frequency spectrum, but because it changes the time structure of the sound. That's temporal masking. And for the next episode, we're going to stop talking about it abstractly. We're going to listen to it.

Practical takeaways

1. Think of masking as a perceptual reduction in audibility, not simply frequency overlap.
2. Critical bands explain why the auditory system does not resolve frequency as infinitely narrow independent lines.
3. Relative level matters, but masking is not simply "the louder sound wins."
4. Musical sounds contain fundamentals, harmonics and transient information, so masking can involve many components simultaneously.
5. Low-frequency maskers can contribute to upward spread of masking.
6. Evaluate masking in the full mix, not only in solo.
7. Identify what information disappears rather than simply looking for spectral overlap.
8. Use context as a diagnostic clue; a source that disappears only when another source enters may be experiencing an interaction problem.
9. Masking can sometimes be desirable because overlap can create fusion, cohesion, weight and size.
10. Don't automatically solve masking with EQ.
11. Consider level, timing, dynamics, register, spatial position, depth and arrangement before assuming EQ is the answer.
12. Use analysers to investigate what you hear, not to decide what you should hear.
13. Don't make rigid rules from critical-band theory.
14. The most useful question is: what information is being lost, and what relationship is causing it?

Episode summary

Why Sounds Mask Each Other explains frequency masking through the psychoacoustic concepts of critical bands and upward spread of masking, while keeping the discussion grounded in practical mixing. The episode distinguishes spectral overlap from meaningful perceptual masking. Musical sounds naturally share frequency regions, and this overlap can create fusion, cohesion and weight. Masking becomes a mixing problem when important information becomes less audible or less intelligible.

Critical bands explain why nearby spectral components can interact within the auditory system. Upward spread of masking explains why strong low-frequency energy can influence the audibility of higher-frequency information. The episode also explains why spectrum analysers are useful diagnostic tools but cannot determine masking by themselves.

The practical framework: hear the problem first, identify what information disappeared, then investigate the relationship that caused it.