Caribbean

Why Folk Keeps Recording the Same Songs

In most genres a cover is a tribute. In folk it is the normal way the repertoire stays alive.

Hand a folk singer a song that has been recorded four hundred times and they will record it again, and nobody in the tradition finds this odd. The assumption that a song belongs to whoever released it first is a recent, commercial idea.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Covering Is How the Repertoire Lives

In most genres recording someone else’s song is a tribute, a curiosity or a chart calculation. In folk it is the normal mechanism by which the material continues to exist.

The tradition is built on songs that have no single authoritative version and frequently no identifiable author, transmitted through performance rather than through publication.

Under those conditions recording a song that others have recorded is not derivative. It is participation in the process that keeps the song in circulation, and it is what the form has always done.

Variation Is the Content

Because there is no definitive version, the interest lies in the differences between performances rather than in fidelity to an original.

Verses appear and disappear, melodies shift between related tunes, and the same narrative turns up with different names, places and outcomes. Collectors documented multiple variants of single songs across regions and generations.

So a new recording is read against the versions a listener already knows. The choices — which verses to include, how to phrase the melody, what to do with the ending — carry the meaning, and they are only legible because the song is familiar.

Regional variants also carry information about transmission routes. The same ballad appearing in two distant places with consistent differences maps how people and songs moved, which is why collectors recorded location and singer alongside the text.

Copyright Sits Awkwardly Over It

The legal framework was designed for identifiable authors and fixed works, which fits this material badly.

Traditional songs are generally in the public domain, so the composition is free to record. What is not free is a specific arrangement, and arrangements of traditional material can be copyrighted by the person who made them.

That has produced genuine disputes, where a performer learns a version from a recording and finds that the particular arrangement is owned. The song is common property; the shape someone gave it may not be.

Recording Changed Transmission

The oral process the tradition depends on was substantially altered by recording, and the effect was double-edged.

Recording preserved a great deal that would otherwise have vanished, and it gave performers access to material from regions they would never have visited. The revival movements depended entirely on archives.

It also fixed versions in a way oral transmission never did. When thousands of people learn a song from the same recording, variation narrows sharply, and a form that survived through mutation began reproducing more exactly.

Copyright on arrangements has an odd side effect: it incentivises performers to alter a traditional song enough to claim an arrangement, which introduces change for legal rather than musical reasons.

The Standard for Judging a Version

Within the tradition, the criteria for a good recording of a known song are reasonably consistent and quite different from those applied elsewhere.

A version is judged on whether the performer has understood the song and found something in it, not on originality of material or technical display. Restraint is usually valued over elaboration, since the song is expected to carry itself.

Adding a great deal of arrangement to a traditional song is not forbidden and is viewed with some suspicion, because the risk is that the production becomes the content and the song becomes an occasion for it.

Why the Practice Persists

The tradition continues to work this way because the alternative would break it.

A repertoire that only survives through performance stops existing the moment performers stop recording it. Every version is both an interpretation and an act of maintenance, and the two cannot be separated.

That is why folk performers show no anxiety about recording well-known material, and why an audience does not experience it as repetition. Each version is another link in a chain, and the chain is the point rather than any individual link in it.

The practice also keeps the repertoire audible to each generation in its own idiom. A song recorded once in 1950 and never again would survive as an artefact; recorded every decade, it stays a living piece of music.

Mk.gee performing live with a Fender guitar

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka performing live with a Fender guitar under purple stage lighting

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro photographed seated on a staircase

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Nilüfer Yanya performing live with her band under stage lighting

Every Nilüfer Yanya Record Runs Through the Same Second Person

Fidelity album artwork, a high-contrast halftone portrait of Yaya Bey

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Arlo Parks performing live under stage lighting on the Ambiguous Desire era

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

How a Pop Vocal Actually Gets Built

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Production & Producers Pop Queer Music & LGBTQA+ R&B Rock

Jazz and the Long Take

Editing an improvised solo is possible and mostly avoided. The reason is structural, not sentimental.

Jazz records could be assembled the way pop records are, from the best fragments of many passes. They largely are not, and the reason has less to do with purism than with what a solo actually is.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Editing a Solo Is Possible and Mostly Avoided

Editing an improvised jazz solo is technically straightforward and has been for decades, and the practice remains rare in a way that requires explanation.

The tools are the same ones that assemble a pop vocal from twenty takes. Nothing prevents a producer from comping the best four bars from each of six solos into a flawless one.

That it is generally not done is not sentimentality about authenticity. It reflects something structural about what an improvised solo is and why editing damages it in a way it does not damage a composed part.

A Solo Is an Argument, Not a Sequence

An improvised solo develops. A phrase introduces an idea, the next responds to it, tension accumulates across choruses, and the resolution depends on everything that came before.

That means the value of any given bar is largely a function of its position in the sequence. A brilliant phrase from take four placed after a phrase from take one is no longer answering anything, because the thing it was answering is not there.

Splicing produces a solo composed entirely of good moments that does not go anywhere, and experienced listeners identify this readily even when the join itself is inaudible.

Take selection in jazz is also frequently made on feel rather than accuracy. A take with a fluffed entry and superior interaction is routinely chosen over a clean one, which inverts the priority applied in most other genres.

The Rhythm Section Is Responding Too

The second obstacle is that a solo is not a solo. The rhythm section is reacting continuously to what the soloist plays, and their reactions are recorded in the same takes.

Editing the soloist without editing everyone else produces a passage where the drummer is responding to something that no longer happens. The interaction breaks even though every individual part is intact.

Editing everyone together preserves the interaction and reintroduces the first problem, since the join now has to work simultaneously across several instruments that were all doing something different at that moment.

Leakage Makes It Physically Hard

There is a practical layer as well. Jazz is frequently recorded live in one room, so every microphone contains every instrument.

That means there is no isolated solo track to edit. Cutting the trumpet channel cuts the trumpet as heard through the piano microphones too, and any mismatch in room sound or ambience at the edit point is audible across the whole mix.

Where a session was recorded with separation, editing becomes more feasible, which is one reason it appears more often on studio-separated fusion records than on live-tracked acoustic ones.

Room microphones make partial editing harder still. Even where instruments were separated, the ambience track contains everything, and a splice in that channel is audible as a shift in the space itself.

What Does Get Edited

It would be inaccurate to suggest jazz records are unedited. Editing happens, and it is generally structural rather than internal.

Whole takes are chosen from several. Choruses are removed to shorten a long solo. Heads and endings are assembled from different takes. An entire solo may be replaced with one from another take of the same tune.

Those operations preserve the internal continuity of any given improvised passage while adjusting what surrounds it, which is a meaningfully different intervention from rebuilding the improvisation itself.

The Value Is in the Document

Underneath the practical arguments is a claim about what the recording is for, and it is worth stating plainly.

A jazz record is understood as documenting a performance — a specific occasion where particular musicians played together and made decisions in real time under some pressure. The interest is partly in what they did and partly in the fact that they did it then.

An assembled solo is a different object. It may be more consistently excellent and it no longer documents anything, which removes the thing the form was recording in the first place.

The convention also sets listener expectations that matter commercially. An audience buying a jazz record believes it is hearing musicians respond to each other in real time, and a record that turned out to be assembled would face criticism a pop record would not.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

How a Pop Vocal Actually Gets Built

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Streaming & Distribution Pop Queer Music & LGBTQA+ R&B Rock

What Release Day Means When Nothing Is Released

The record does not ship, arrive or sell out. Release day survives because the industry needs a coordination point.

Nothing physically happens on release day. A file that was already sitting on a server becomes visible. The date persists anyway, and it does real work.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Nothing Actually Ships

Release day is a fixed, heavily promoted moment in a process where nothing physical happens. No record arrives at a shop, no stock is unpacked, and nothing sells out.

The term is inherited from an era when it described a genuine logistical event: pressed records reaching retailers, who could not sell them earlier and were expected to sell them from that morning.

For a digital release the equivalent is a file becoming visible, which could occur at any moment and requires no coordination with anyone. Yet the convention has strengthened rather than weakened.

It Survives Because Attention Needs Coordinating

The reason it persists is that the industry needs a point around which everything else can be synchronised, and that need has grown rather than diminished.

Playlist consideration, press embargoes, radio adds, video premieres, social campaigns, advertising and retail features all have to be aimed somewhere. Without a fixed date they cannot reinforce one another.

The date is therefore not about distribution at all. It is a coordination device that lets a large number of independent parties concentrate their activity into the same window.

Embargo coordination is the practical work the date makes possible. Press, radio and playlist teams all need to act simultaneously, and without a shared moment each would publish whenever ready, dissipating the attention rather than concentrating it.

Chart Weeks Impose the Structure

There is also a mechanical reason the specific day matters, which is that charts measure fixed periods.

A release landing at the start of a chart week accumulates a full week of activity toward its first-week position; one landing near the end gets a truncated window. That difference is large enough to change a chart placing substantially.

The industry converged on a common global release day partly to serve this and partly to prevent territory-by-territory staggering, which had become unworkable once anyone could see what had been released elsewhere.

The Cliff Is a Genuine Problem

The convention produces a well-known and awkward pattern: a sharp spike of attention on the day, followed by a rapid decline over the following fortnight.

That suits the chart and suits playlist refresh cycles, and it fits how records actually find audiences rather badly, since discovery is continuous and a record’s audience typically builds slowly.

Artists and labels have responded by extending campaigns — releasing singles over months, staggering videos, adding deluxe editions later — all of which are attempts to create additional moments because the single moment is insufficient.

The cliff also shapes what gets made. Records designed to peak in week one are structured differently from records designed to accumulate, and the convention quietly favours the former.

Alternatives Exist and Are Marginal

Surprise releases demonstrate that the convention is not technically necessary, and they only work under specific conditions.

An artist with a very large existing audience can generate their own coordination point by announcing a record that is already available. Attention arrives because the artist is the event, not because a schedule aligned.

For everyone else, forgoing the coordination means forgoing playlist consideration, press and radio, all of which require lead time. The convention persists because the alternative is available only to people who do not need it.

A Vestigial Structure That Still Works

The honest description is that release day is a convention inherited from physical distribution, kept alive by the coordination requirements of promotion and by the arithmetic of chart weeks.

It no longer describes anything happening in the world, and it continues to do useful work, which is a reasonable definition of a functional institution rather than a broken one.

What has changed is that it is now one moment in a longer campaign rather than the campaign itself, and treating it as the finish line rather than a milestone is the mistake most commonly made by artists releasing for the first time.

The strongest argument for keeping it is that a shared date is what allows anything to be an event at all. Continuous release means nothing ever arrives, and audiences respond to arrival.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

How a Pop Vocal Actually Gets Built

By Will Lisil / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Production & Producers Pop Queer Music & LGBTQA+ R&B Rock

The Click Track Problem in Orchestral Recording

Playing to a metronome is standard for film scores and rare for concert recordings. The difference explains a lot.

Ask an orchestra to play to a click and you get precision. You also get something subtly different from what the same orchestra does without one, and both facts matter depending on what the recording is for.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Standard for Film, Rare for Concert Work

Playing to a click track is routine in film scoring and unusual in concert-hall recording, and the difference explains a great deal about how each sounds.

Film music has to align with picture at specified moments, which means the tempo has to be predictable and repeatable. A click makes synchronisation possible and makes later editing straightforward.

Concert repertoire has no such requirement, and orchestras recording symphonic works generally follow a conductor with no reference tempo at all. The result is audibly different in ways that go beyond precision.

Orchestras Do Not Play at a Constant Tempo

The central issue is that expressive orchestral performance involves continuous small tempo variation, and that variation is not incidental decoration.

Phrases are shaped by pushing forward and holding back. Cadences broaden. Transitions between sections accelerate or relax. Those movements are how the music communicates structure, and they are written into the tradition rather than added by individual conductors.

A click removes all of it by definition. Playing accurately to a fixed pulse means suppressing exactly the flexibility that expressive playing consists of, which is why click-recorded orchestral music can sound technically excellent and emotionally flat.

Musicians also report that a click changes their relationship to each other. Following a shared external pulse is a different cognitive task from listening to the person beside you, and ensembles describe the result as accurate but less connected.

Ensemble Coordination Works Differently

There is a practical problem beneath the aesthetic one, which is that a large ensemble does not synchronise to a metronome the way a small group does.

Players in an orchestra coordinate visually with a conductor and aurally with the sections around them, and there are genuine physical delays across a large stage. Sound takes measurable time to travel from one side to the other.

A click delivered through an earpiece gives each player an identical reference that ignores those delays, which can actually degrade ensemble rather than improve it, particularly in the back desks.

Film Sessions Solve It With Variable Maps

Film composers work around this rather than accepting a rigid pulse, and the solution is well established.

Rather than a constant tempo, sessions use a tempo map with programmed variation built in — accelerations, ritardandos and fermatas written into the click itself. The orchestra still plays to a reference, and the reference breathes.

Streamers and punches, visual cues projected for the conductor, do similar work by marking hit points without imposing a pulse between them. Both approaches try to recover flexibility while keeping synchronisation.

Punches and streamers remain in use precisely because they give the conductor timing information without imposing a pulse on the players, which is the distinction that makes them tolerable where a click is not.

Editing Is the Real Driver

The reason click use persists in film work is not synchronisation alone. It is that the material must remain editable after the session.

Picture changes constantly, and music recorded to a fixed grid can be cut, looped and repositioned to fit a new edit. Music recorded with free tempo cannot be manipulated that way without becoming obvious.

That flexibility is worth a great deal on a production where the edit is not locked, and it is the practical reason composers accept a constraint they would not otherwise choose.

Let the Requirement Pick the Method

The reasonable position is that neither approach is correct in general and that the choice should follow the requirement.

Material that must synchronise to picture and survive re-editing needs a click, ideally with a variable tempo map that preserves as much flexibility as the deadline allows.

Material recorded as a performance, intended to stand on its own, is better served by a conductor and no reference at all. Applying film-session methods to concert repertoire imports a constraint that solves a problem the recording does not have.

Sample-based mock-ups have made the problem worse in one respect. Directors approve a gridded demo, then find the live orchestra’s flexible version no longer matches the edit they have been watching for months.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

How a Pop Vocal Actually Gets Built

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Discussion & Analysis Pop Queer Music & LGBTQA+ R&B Rock

The Return of the Arrangement in R&B

After years of sparse, loop-led production, a lot of R&B is being written with parts again.

For most of the last decade the dominant R&B production style was subtraction: a loop, a sparse drum pattern, a great deal of space. A noticeable amount of current R&B is doing the opposite, and writing arrangements again.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Parts Are Returning

After an extended period in which R&B production was dominated by sparse, loop-led arrangements, a significant amount of the genre is being written with distinct instrumental parts again.

The loop-led approach had real advantages: it was fast, it left enormous space for a vocal, and it suited a production process built around producers sending beats rather than bands working in rooms.

What is appearing now is closer to arrangement in the older sense — bass lines that move independently, chord voicings that change across a section, horn and string writing, and drum parts that develop rather than repeat.

Live Musicians Came Back First

The shift is traceable partly to a return of played instruments, particularly bass and keys, on records that would previously have used programmed parts.

A played bass line does not repeat identically, and it responds to the harmony rather than sitting under it. That single change forces the rest of the arrangement to accommodate movement it previously did not have.

Once one part moves, others tend to follow, because a static pad underneath a moving bass sounds increasingly like an omission rather than a choice.

Arranger credits have returned alongside the parts, which is a meaningful change in how these records are made. A named arranger implies written charts and a session rather than a producer assembling parts alone.

Harmonic Vocabulary Widened Again

Alongside that, the harmonic language has become more elaborate, drawing again on the extended chords and voice-leading that connected earlier R&B to jazz.

Loop-led production tends toward a repeating two- or four-bar harmonic cell, which constrains how much a progression can develop. Writing with parts permits longer harmonic structures with genuine movement and resolution.

That in turn changes what vocalists can do. A melody written over a developing progression can go somewhere that a melody over a repeating cell cannot, and vocal arrangements have become correspondingly more ambitious.

The Vocal Stopped Carrying Everything

In the sparse era, the vocal carried nearly all the musical information, with the production functioning as a frame around it.

That produced some exceptional vocal performances and placed considerable weight on a single element. When the arrangement contributes, the vocal can do less and still be effective, which opens up restraint as an option.

Records in this mode tend to use the voice as one part among several rather than as the entire foreground, which is closer to how the genre worked historically.

Recording a live rhythm section also changes the vocal approach. Singing against players who respond invites a different, more conversational performance than singing to a fixed loop, and the vocals on these records are audibly less gridded.

It Costs More and Takes Longer

The reason the sparse approach dominated was not only aesthetic. It was substantially cheaper and faster, and the return to arrangement reverses both.

Written parts require players, sessions, someone to write them, and time to record and mix a larger number of elements. A record with strings and horns is a materially different budget from one built on a loop and a vocal.

That means the shift is uneven, appearing first among artists with resources and among independents who prioritise it, rather than as a wholesale change across the genre.

A Correction Rather Than a Rejection

The most accurate description is a correction rather than a repudiation of what came before.

The sparse, loop-led style produced a substantial body of good records and remains entirely viable. What changed is that it stopped being the only available register, so its use became a choice again rather than a default.

That is generally how these cycles resolve. A dominant approach becomes the baseline, an alternative develops in response, and the genre ends up with both available — which is a better position than it occupied at either extreme.

The shift is uneven by budget rather than by taste, which is worth stating. Written arrangements and hired players cost money, so the correction is visible first among established artists and well-funded independents.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

Stem Separation Changed What a Remix Is

How a Pop Vocal Actually Gets Built

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Technology & Innovation Pop Queer Music & LGBTQA+ R&B Rock

Stem Separation Changed What a Remix Is

Pulling a clean vocal out of a finished mix used to be impossible. Now it takes seconds, and the consequences are still being worked out.

For most of recorded history, a finished mix was final. The individual parts were welded together and could not be recovered. Machine-learning separation has undone that, and the implications reach a long way past remixing.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Extraction Became Trivial

Pulling a clean vocal out of a finished stereo mix was effectively impossible for most of recorded history, and it now takes a few seconds on a laptop.

Machine-learning separation models, trained on very large quantities of music with known component parts, can distinguish vocals, drums, bass and other instruments within a mixed recording and output them separately.

The quality is good enough for practical use. Artefacts are audible under scrutiny, particularly on dense material, and for remixing, sampling and live performance the results are entirely usable.

It Changed Who Can Remix

The immediate consequence is that access to official stems stopped being a prerequisite for remixing.

Historically a remixer needed the rights holder to supply the multitrack, which meant remixing was something arranged through labels and granted to established producers. That gatekeeping is gone.

Anyone can now extract a vocal from any recording and build something around it, which has produced an enormous volume of unofficial remixes and a corresponding rights problem that nobody has resolved cleanly.

Separation quality also varies sharply by source material. Sparse recordings separate cleanly; dense, heavily compressed mixes produce audible artefacts, and material where instruments share a frequency range confuses the models most.

The Legal Position Is Unchanged

It is worth being clear that the technology has not altered the law, only the practical difficulty of infringing it.

An extracted vocal is a derivative of a copyrighted master. Using it without permission is infringement in the same way sampling without clearance is, regardless of how the extraction was performed.

What has changed is enforcement reality. Platforms detect and act on some of this automatically, a great deal circulates without challenge, and rights holders have generally chosen to tolerate promotional uses while pursuing commercial ones.

Archival and Restoration Benefits Are Real

The most straightforwardly positive application is on material where the multitracks no longer exist, which is a large proportion of recorded history.

Separation permits genuine remixing of recordings that survive only as a stereo master, allowing balance corrections, noise reduction targeted at one element, and surround or spatial versions that were previously impossible.

Several high-profile catalogue projects have used exactly this, and the results have been accepted by audiences who would have rejected a conventional remaster of the same material.

Archival applications have an ethical dimension too. Remixing a recording whose artists are dead means making balance decisions they never approved, which is a real objection even where the technical result is superior.

DJs and Live Performers Use It Constantly

A quieter but larger use is in performance, where separation has become routine rather than exceptional.

Removing a vocal to create an instrumental for mixing, isolating a drum part to layer under another track, or extracting an acapella for a live mashup are now standard practice, executed in real time by software built into performance systems.

That has changed what a DJ set can contain, since any recording is now effectively available in component form regardless of what the label released.

The Unresolved Question Is Consent

What remains genuinely unsettled is whether artists have any say in their work being decomposed this way, and currently they largely do not.

A vocal performance can be lifted from its arrangement and placed in a context the singer would have refused, and the technical barrier that previously made this difficult has gone. Existing law addresses reproduction rather than this kind of recontextualising.

Some rights holders have begun releasing official stems to get ahead of it, on the reasoning that supplying the material lets them set terms. That is probably the direction of travel, since the alternative is attempting to prohibit something that anyone can already do.

Official stem releases are probably where this settles. Supplying the parts lets a rights holder set terms and formats rather than trying to prohibit something any listener can now do in seconds.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

How a Pop Vocal Actually Gets Built

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Production & Producers Pop Queer Music & LGBTQA+ R&B Rock

How a Pop Vocal Actually Gets Built

The lead vocal you hear is almost never one performance. Understanding how it is assembled changes how you hear the genre.

The single most misunderstood thing about modern pop is the lead vocal. It sounds like one person singing once. It is almost never that.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

The Lead Is a Composite

The lead vocal on a contemporary pop record is almost never a single performance, and understanding how it is assembled changes how you hear the genre.

What reaches the listener is a composite, built from a large number of takes with the best line, phrase or sometimes syllable selected from each. Engineers call the process comping, and it is standard practice rather than a remedy for a weak singer.

A confident vocalist may sing a song twenty times. The comp then takes the first verse from take four, the pre-chorus from take eleven, and a single word from take nineteen because that is where the consonant landed correctly.

Comping Is an Editorial Job

Assembling the comp is closer to editing prose than to engineering, and it is where a great deal of a record’s character is decided.

The person doing it is choosing between performances that differ in pitch, timing, tone, attitude and breath, and those choices determine whether the finished vocal sounds urgent, weary, casual or precise. Two engineers comping the same session produce noticeably different records.

The skill is in maintaining continuity. Lines assembled from takes recorded an hour apart have to sound like one continuous thought, which means matching not just pitch but energy and the way each phrase leads into the next.

Comping is also where a producer’s relationship with a singer is tested. Choosing a take that is technically weaker but emotionally right means telling a vocalist that their most controlled performance is not the one being used, and that conversation goes better when it happens during the session rather than afterwards.

Doubles and Stacks Build the Body

Beyond the lead, most pop vocals include additional layers whose purpose is thickness rather than melody.

A double is the same line sung again and placed alongside the lead, which reinforces it and smooths small imperfections. Stacks are multiple harmony parts, often several per note, sometimes recorded many times to create a choir from one singer.

These are frequently used selectively rather than throughout. A chorus might have eight layers while the verse has one, which is a large part of why choruses feel like they open up. The arrangement is happening in the vocal itself.

Tuning Is Applied to Almost Everything

Pitch correction is close to universal on commercial pop vocals, and its most common use is invisible rather than an effect.

Applied gently, it moves notes toward their target without removing the natural pitch movement of the voice. The result sounds like a singer with excellent intonation rather than like a processed vocal, which is the intention.

It matters more on stacked parts than on the lead, because small pitch differences between layers that would be pleasant on one voice become audible beating when eight are stacked. Tuning the stack is what makes it sound like a wall rather than a crowd.

Tuning stacks is a different job from tuning a lead. Small pitch differences between layers that sound like richness on two voices become audible beating on eight, so stacked parts are corrected harder than the lead they support, which is why the lead retains character the stack does not.

Compression Does the Consistency Work

A voice has an enormous dynamic range, and pop production requires it to sit at a consistent level against a dense backing, which means substantial dynamic control.

This is usually applied in stages rather than all at once: a first compressor evening out the largest variations, a second adding density, and often manual level adjustment written in before either. Doing it in stages keeps any single processor from working hard enough to become audible.

Manual gain riding before compression is the step most often skipped and the one that most improves the result, because it lets the compressors handle character rather than fighting the performance.

Why This Matters to a Listener

None of this is concealed within the industry, and knowing it changes what you are listening to in a useful way.

A pop vocal is a constructed object presenting an idealised performance, in the same way a film performance is constructed from takes and coverage. Judging it as a document of one person singing once is applying the wrong standard.

It also reframes the skill involved. The singer has to deliver twenty usable performances with consistent intent, and someone else has to assemble them into something that sounds inevitable. Both are considerable, and neither is what the finished record appears to show.

Knowing this also reframes live performance. A singer reproducing a comped, stacked, tuned studio vocal alone on a stage is attempting something the record never required of anyone, which is worth remembering before judging a live vocal against its recorded version.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Production & Producers Pop Queer Music & LGBTQA+ R&B Rock

Why Guitar Bands Went Back to Tracking Live

Recording a band together in one room is slower, more expensive and harder to fix. A lot of groups are choosing it anyway.

For two decades the default way to record a rock band was one instrument at a time. A growing number of groups have gone back to standing in a room and playing at once, and the reasons are more practical than romantic.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

Choosing the Harder Method

Recording a band playing together in one room is slower, more expensive and much harder to fix afterwards, and a substantial number of guitar groups are choosing it anyway.

The dominant alternative — tracking drums first, then bass, then guitars, then vocals — offers total control. Every part can be perfected in isolation and replaced at any point without affecting anything else.

Groups returning to live tracking are giving up that control deliberately, which suggests they are getting something in exchange that the controlled method does not provide.

Interaction Only Happens in Real Time

The thing being bought is ensemble interaction, and it is not available in a layered session because it requires musicians responding to each other as they play.

A drummer pushing slightly ahead into a chorus while the bassist follows, a guitarist leaving space because the singer did something unexpected, a whole band lifting on the last chorus — these are reactions, and they need everyone present.

Layered recording can reproduce the notes but not the reactions. Each player is responding to a fixed recording rather than to a person, and the result tends toward accuracy without the small mutual adjustments that make a band sound like one thing.

Headphone mixes are the practical obstacle most often underestimated. Getting six players each hearing a balance they can perform to takes real time and a monitoring system that many rooms do not have, and a bad headphone mix will ruin takes that the band could otherwise play.

Micro-Timing Carries the Feel

The mechanism underneath this is timing at a scale too fine to notate. Players in an ensemble consistently place notes slightly ahead of or behind the beat relative to each other, and those relationships are stable and characteristic.

That is a large part of what listeners describe as feel or groove, and it emerges from mutual adjustment. When each part is recorded separately against a click, every player aligns to the grid rather than to each other, and the relationships disappear.

Quantised, grid-aligned playing is not wrong, and it suits a great deal of music. It is simply a different thing, and it cannot be converted back into ensemble timing by adding swing afterwards.

Commitment Speeds Everything Up

There is a practical benefit alongside the musical one, which is that live tracking forces decisions that layered recording allows you to postpone indefinitely.

When a take is the take, arrangement questions get settled in the room. Nobody spends three weeks deciding whether the second guitar should be there, because the second guitar was either played or it was not.

Records made this way frequently finish faster overall despite the session being harder, because the mixing stage has far fewer options to evaluate and far fewer decisions left outstanding.

Deciding what constitutes a keeper also changes. In a layered session a take is judged part by part; live, a take with one weak moment and five excellent ones may still be the right choice, and making that call requires someone willing to accept an imperfection.

The Costs Are Real

Being honest about the trade matters. Live tracking requires a room large enough to hold a band with acceptable separation, which is expensive and not widely available.

It requires a band that can play the material properly, since there is no fixing a part that was never performed well. That means rehearsal time, which is its own cost.

And leakage means individual elements cannot be substantially changed later. A guitar tone that turns out to be wrong is wrong on the record, because it is in every microphone.

Why It Is Coming Back Now

The timing is not coincidental. As production tools have made grid-perfect, fully controlled records trivially achievable, that sound has become the default and stopped signalling anything.

A record that audibly documents people playing together is now the distinguishing choice, in the same way that deliberate imperfection has become the marked option in other areas of production.

There is also an audience argument. Listeners cannot usually identify what makes a live-tracked record different, and they respond to it consistently. Whatever the mechanism, the ensemble quality survives compression, small speakers and distracted listening in a way that fine production detail does not.

The revival also tracks the availability of rooms. Large live spaces have become scarcer and more expensive as studios closed, so the bands doing this are frequently working in converted buildings and residential studios rather than in the facilities the method was developed in.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Production & Producers Music Technology & Innovation Pop Queer Music & LGBTQA+ R&B Rock

The Loudness War Ended and Nobody Announced It

Streaming normalisation quietly removed the reward for crushing a master. The habits took longer to fade than the incentive.

For about twenty years, records got louder because loud won. Then the reason evaporated, and it took the industry a surprisingly long time to notice.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

The Incentive Disappeared Quietly

The loudness war ended without an announcement, because what ended it was a technical change to playback rather than a change of heart in mastering.

For decades there was a genuine competitive reason to master loud. A louder record sounded better in direct comparison, and comparisons happened constantly — on radio, on shuffled CD changers, in shops. Being quieter meant sounding weaker.

Streaming loudness normalisation removed that. Platforms now adjust playback level so tracks arrive at broadly similar perceived loudness, which means a heavily limited master is simply turned down to match a dynamic one.

Turning It Down Exposes What Was Lost

The consequence is worse than neutral for the loud master, and this is the part that changed practice.

When a crushed track is attenuated to match a dynamic one, the listener hears it at the same volume but with all the transients gone and no dynamic range. Set beside a well-mastered record at equal loudness, it sounds flat and lifeless rather than powerful.

So the trade that once bought competitive advantage now buys a straightforward disadvantage. The record gave up its dynamics and received nothing in return, which is the clearest possible argument against the practice.

Genre differences are substantial and rarely acknowledged. Material that is dense and mid-heavy suffers most from limiting, while sparse acoustic recordings can absorb a surprising amount before anything audible happens, so a single loudness target was always going to damage some music more than others.

Habits Outlasted the Reason

Despite the incentive vanishing, aggressive limiting continued for years, and the persistence is instructive.

Part of it was that loud masters had become the expected sound, and clients hearing a dynamic master in an unnormalised context judged it as weak. Engineers were asked for loudness by people comparing files on a desktop, where no normalisation applied.

Part was simply established practice. A generation of engineers had learned mastering in an environment where loudness was the objective, and professional habits do not change because a specification did.

Measurement Made the Argument Winnable

What eventually shifted practice was the availability of standard loudness measurement, which turned a subjective argument into a numerical one.

Integrated loudness measured to a common standard gave everyone a shared vocabulary. An engineer could show a client exactly how much the platform would attenuate their master and demonstrate the consequence.

Tools that simulate platform normalisation during mastering made this concrete. Hearing your own loud master turned down against a dynamic reference is more persuasive than any explanation.

Client education turned out to be the decisive factor rather than any technical development. Engineers who could demonstrate the effect on a client’s own material, at matched loudness, won arguments that no amount of general explanation had settled.

Dynamics Are Now a Choice Again

The current situation is that loudness is no longer a competitive variable, which returns dynamic range to being an artistic decision.

Genres where relentless density is part of the point can still be mastered that way, deliberately, and normalisation costs them nothing they were trying to keep. Material where dynamics carry meaning can retain them without being disadvantaged.

That is a considerably healthier arrangement than the one it replaced, where every record was pushed toward the same target regardless of whether it suited the music.

What Has Not Been Recovered

One consequence is permanent and worth stating: a great many records mastered during the loudness era exist only in crushed form.

Where the original mixes survive, remastering can restore dynamics, and a number of albums have been reissued with the limiting removed. Where they do not, the loud master is the record.

That is the lasting cost. The incentive lasted roughly two decades, and the recordings made under it carry the compromise permanently, which is a reasonable argument for treating any current mastering convention with some suspicion.

Vinyl provided an unexpected corrective. A cutting engineer cannot cut an over-limited master cleanly, so records intended for vinyl were frequently prepared from a more dynamic version, and in several cases that version is the better-sounding release.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…

Kae Tempest Sampled His Own Old Lyrics to Argue With Them

Nala Sinephro Scored an A24 Fight Film Without Once Sounding Like One

Every Nilüfer Yanya Record Runs Through the Same Second Person

Yaya Bey Built Fidelity Around Three Deaths and One Miami Hotel Room

Ambiguous Desire Rebuilds Arlo Parks Around a New York Dancefloor

Rachel Chinouriri’s Second Album Is One Night Out, Five Years Long

Why Folk Keeps Recording the Same Songs

Jazz and the Long Take

What Release Day Means When Nothing Is Released

The Click Track Problem in Orchestral Recording

The Return of the Arrangement in R&B

Stem Separation Changed What a Remix Is

By Esther Winslow / August 25, 2026

Afro Music Blues Caribbean Cinematic Classical Country Electronic Folk Hip-Hop Holiday Music Jazz Music Industry Pop Queer Music & LGBTQA+ R&B Rock

What an A&R Actually Does Now

The job used to be finding people in rooms. Most of the finding is done by the time an A&R arrives.

The popular image of A&R is someone in a back room at a small venue, discovering a band nobody has heard. That version of the job has been quietly obsolete for years.

You can listen to our full playlist which contains the artist’s music, and know more about the artist’s work by scrolling down the page.

PLAYLIST

The Finding Happens Before They Arrive

The A&R role used to be built around discovery in physical spaces — being in venues, hearing bands before anyone else, and forming a judgement ahead of the market.

That is largely not how signings originate now. By the time an A&R is seriously interested, an artist usually has streaming numbers, an audience, and often other labels in the conversation.

The discovery has typically been done by the audience, and the data is visible to everyone in the industry simultaneously. What was once an information advantage has become a shared dashboard.

Data Levelled the Field

Streaming analytics, social metrics and playlist performance are broadly available, which means several labels usually identify the same artist at the same time.

That changes the job from finding to winning. Where an A&R once competed on taste and access, they now compete on relationship, terms, and the credibility of what the label will actually do.

It also compresses the timeline. An artist showing sharp growth attracts multiple approaches within weeks, and the process that once took months of observation now happens in a few conversations.

Contract structures shifted alongside the role. Where a label once funded development over several records, deals now more frequently attach to an artist who has already demonstrated an audience, which changes the risk profile and the terms an artist can negotiate.

Development Moved to the Artist

The other substantial change is that artists arrive far more finished than they used to.

A signed act today typically has recorded material, a visual identity, an audience and a working knowledge of how to release. Much of the development the label once provided has already happened, funded by the artist.

That reduces the label’s contribution at the early stage and shifts it later, toward scale: radio, international expansion, large-scale marketing and the infrastructure an independent artist cannot replicate.

The Job Is Now Largely Internal

A significant part of contemporary A&R work happens inside the label rather than outside it, and it is not what the title suggests.

Securing budget, marketing attention and priority for an artist against everyone else on the roster is a continuous internal negotiation. A signed act with no institutional advocate receives little, regardless of the deal terms.

That makes the A&R the artist’s representative within the company as much as the company’s representative to the artist, and it is the part of the role least visible from outside.

The internal advocacy is also why the identity of the signing A&R matters to an artist’s outcome. If that person leaves the company, the artist frequently loses their route to budget and attention, which is a well-known and rarely disclosed risk of signing.

Judgement Still Matters, Differently

None of this means taste has become irrelevant, but what it is applied to has changed.

The question is no longer whether an artist is good, since the numbers indicate that an audience already thinks so. It is whether the growth will continue, whether the artist can produce a second and third record, and whether they can withstand the process.

Those are judgements about people and durability rather than about music, and they are harder to make from data. This is where experienced A&Rs still outperform, and it is why the role has not been automated.

A Narrower Role, Still Necessary

The honest summary is that the job has contracted in scope while remaining consequential.

Discovery is largely gone, early development is substantially gone, and the information advantage that defined the role has been distributed. What remains is relationship-building, internal advocacy and long-horizon judgement about people.

That is a real job, and it is not the romantic one the title still carries. The A&R who found a band in a basement before anyone else is now mostly a story the industry tells about itself.

What has not changed is that the decision is ultimately about a person rather than a dataset. Whether someone can produce a second record under pressure is not visible in streaming analytics, and getting that judgement right remains the whole job.

Mk.gee Made a 33-Minute Debut With No Guests and No Clear Vocals

Will Lisil

August 28, 2026

Almost every instinct in modern record-making pushes the vocal forward. Streaming rewards an identifiable hook early, playlists reward clarity, and…

Michael Kiwanuka Made Small Changes by Not Trying to Prove Anything

Will Lisil

August 28, 2026

There is a particular problem that arrives after an album wins something. The next record has to answer it, and…