Production · 2026-09-13 · 5 min read

What Is the Difference Between a Producer, a Mixing Engineer, and a Mastering Engineer?

Three roles, three decisions, three handoffs. What each person owns, the file that passes between them, and why one name sometimes fills all three.

The producer decides what the record is, the mixing engineer decides how its parts sit together, and the mastering engineer decides how the finished stereo file behaves on every system that plays it. Each role owns one decision and hands the next a specific file. The producer works in the arrangement and the performance, the mixing engineer works inside the multitrack, and the mastering engineer works on the final two-track. I hold all three roles on some records and only one on others, and the line between them is the handoff, not the title. Here is what each stage decides and what travels to the next.

What does the producer decide?

The producer owns the version of the song that gets recorded. Tempo, key, arrangement, which take, how long the intro runs, whether the second verse drops the drums, whether the hook needs a counter-line. On a pop record aimed at phones I will cut an eight bar intro to four so the hook arrives before a listener swipes, and that single call changes what everyone downstream is handed.

The producer also owns the sound at the source: the microphone on the vocal, the room, the synth patch, the comp of the best syllables across ten passes. By the time production is finished the song has a shape and a rough balance. On "Stefania" the Qobuz metadata from Sony Music lists the same name as composer and producer, so the arrangement decisions and the writing decisions came from one desk. Those credits sit at /credits.

Where does producing end and mixing begin?

At the handoff. The producer passes the mixing engineer an organized session: tracks labeled, edits committed, a rough balance, and a note on what the vocal should feel like against the track. Gain staging matters here more than most people expect, so I leave individual tracks averaging near -18 dBFS with peaks well under clipping, which gives the mix stage headroom to add rather than fight.

The mixing engineer then owns how the parts relate. Balance, panning, front to back depth, the EQ moves that keep the bass and the kick out of each other's way, the compression that makes a vocal sit at one level without sounding squashed, the reverb and delay that place each part in a space. The producer decided there is a tambourine. The mixing engineer decides you can hear it in the second chorus and not the first.

What does the mixing engineer hand the mastering engineer?

One stereo file, usually 24-bit at the session rate, with peaks around -6 dBFS and no limiter strapped across the two-bus. Stems come too if the master might need to nudge the vocal or the low end. The reason for the headroom is simple: the mastering engineer needs level to work with, and a mix already crushed to the ceiling leaves nothing to shape. I would rather deliver a mix that sounds slightly quiet on its own and masters well than one that is loud and locked.

A mix printed with a limiter already doing three decibels of work has made a mastering decision inside the mix, and it cannot be taken back later.

What does mastering do that mixing does not?

Mastering works on the finished two-track and on the record as a whole. It sets the final tonal balance so a whole EP sounds like one release, it controls loudness against a target, and on an album it sequences the songs and sets the gaps between them. The numbers are concrete: most streaming delivery aims for about -14 LUFS integrated with true peak at or below -1 dBTP (Sound On Sound on loudness for streaming), while a vinyl cut or a CD master needs different handling, and a CD master gets dithered down to 16-bit. The trade between loudness and dynamics is the whole craft, which I wrote about in vinyl mastering against a streaming master.

Mastering is also the last quality gate. It catches the click at the top of the file, the sibilant that survived the mix, the low rumble under the intro that no one heard on small speakers.

When is it the same person, and when should it not be?

On a small release, one person carrying all three roles is faster and keeps the intent coherent from arrangement to master. On major-label pop the roles are split so each stage gets fresh ears, and mastering benefits most from that distance, because the person who made the mix decisions is the worst judge of whether they worked.

Both models are real in my own credits. On "Stefania", which won the 2022 Eurovision Song Contest), the same name held the writing and engineering roles. On "In The Shadows Of Ukraine", Kalush Orchestra's release with the Finnish band The Rasmus on Columbia, I produced and mastered the record without mixing it (NME on the release). A split can put the strongest person on each stage. One pair of hands can keep a small record moving.

Why do these credits keep going missing?

Because streaming display strips them. The producer, the mixing engineer and the mastering engineer are separate lines in the release metadata, but most players show only the artist and maybe the label. The fix lives in the delivery metadata, not the app, which I set out in why a producer credit disappears on streaming. Knowing which role did what is also how you hire: a song with a great arrangement and a weak master needs a different call than a great master on a thin mix.

Frequently asked

Can one person be producer, mixing engineer and mastering engineer on the same song? Yes. On "Stefania" the Qobuz metadata lists the same name across the writing and engineering roles. It is common on smaller releases and rarer on major-label pop, where the jobs are usually split for a second set of ears.

What does a mixing engineer hand to a mastering engineer? Usually one stereo file at 24-bit, peaks near -6 dBFS, no limiter on the two-bus, plus stems if the master might rebalance. The headroom is the point, because mastering needs room to work.

What loudness should a master target for streaming? Streaming platforms normalize toward roughly -14 LUFS integrated and ask for true peaks at or below -1 dBTP. A master pushed far louder gets turned down on playback and loses dynamics for nothing.

If you are not sure which stage you need

Send the song and one line on what is wrong: the arrangement, the balance, or the way the finished file translates. What comes back is a plain answer on whether it is a production, a mix, or a master problem, and what the next file should look like. Background and contact at /about.

I(Questions)

Can one person be producer, mixing engineer and mastering engineer on the same song?

Yes. On "Stefania" the Qobuz metadata lists the same name across the writing and engineering roles. It is common on smaller releases and rarer on major-label pop, where the jobs are usually split for a second set of ears.

What does a mixing engineer hand to a mastering engineer?

Usually one stereo file at 24-bit, peaks near -6 dBFS, no limiter on the two-bus, plus stems if the master might rebalance. The headroom is the point, because mastering needs room to work.

What loudness should a master target for streaming?

Streaming platforms normalize toward roughly -14 LUFS integrated and ask for true peaks at or below -1 dBTP. A master pushed far louder gets turned down on playback and loses dynamics for nothing.

Need this on your record?

Start a session →