Mixing and Mastering: What Each One Actually Does to Your Song

Mixing takes the individual recorded tracks of your song — vocal, percussion, bass, strings, keys — and turns them into one balanced stereo picture. I set levels, place instruments left and right, shape each sound with EQ and compression, and add the depth: reverb, delay, saturation.

The tool I lean on more than any effect is automation. A good mix moves constantly. The vocal comes up half a decibel for a quiet word, the strings dip when the singer enters, the reverb blooms at the end of a phrase and dries up before the next one starts. That invisible, ongoing adjustment is the difference between a mix and a set of faders left in one position.

If a lyric is hard to hear, if the chorus doesn't lift, if the low end turns to mud in the car — those are mixing problems, and they can only be fixed inside the individual tracks.

Mastering happens after the mix is approved. I work on the finished stereo file, not the separate tracks: final EQ decisions, gentle compression, the final loudness for release, and a check that the song translates — that it sounds right on earbuds, in a car, on a club system, and on a phone speaker.

Here's a way to keep them straight: mixing decides how the instruments relate to each other; mastering decides how the finished song relates to every other song in the world. In mastering I can make the whole record warmer, brighter or punchier — but I can no longer turn the derbake down on its own. If the balance is wrong, I go back into the mix.

That's why I treat mixing and mastering as one process with two stages. The second stage can only polish what the first stage got right.

Arabic vocal mixing is its own craft. The lead is ornamented — a singer can pass through six or seven notes on a single syllable — and the arrangement underneath is dense: derbake, riq and daf filling the midrange, string lines doubling the melody. Sitting that voice on top of all of it, without making it shout, is most of the work.

Tuning has to be done by hand, note by note. Automatic pitch correction snaps everything to the Western scale, which flattens the ornaments and erases the quarter tones that many maqams are built on. I correct the note the singer aimed for — including the notes that sit between the piano keys — and I leave the slides and turns alone. An Arabic ear catches an algorithm-tuned vocal within two seconds, even if the listener can't say why.

A mawwal needs its own treatment. It's free time over sparse backing, so the voice is completely exposed: I give it a longer, wider reverb than the rest of the song and ride the level word by word. The moment the rhythm drops in, that space has to tighten instantly — the contrast between the open mawwal and the locked groove is an effect in itself, and it's built in the mix.

Arabic pop loves big vocal reverb, and I don't fight the genre. The craft is keeping the words clear inside it: pre-delay so the reverb starts after the word instead of on top of it, EQ that keeps the tail out of the consonant range, and ducking so the reverb blooms in the gaps between phrases rather than washing over them.

Against all that, the percussion has to keep its attack and the strings their weight. I'd rather automate the strings down a touch under every vocal phrase than thin them out with EQ — the listener should feel a full arrangement and still catch every word.

Stems are the individual tracks of your session, each exported as its own audio file. Here's how to prepare stems for mixing so nothing gets lost between your studio and mine: If you are earlier than this — no session yet, just an idea in your head — start with how an Arabic pop song actually gets made instead.

When I send the first mix, listen on at least two systems — your usual headphones and a car or phone speaker — and live with it for a day before writing notes. Then send everything in one batch, with timestamps: "the vocal at 2:14 feels buried", "less reverb on the second verse". Notes like that I can act on precisely. "Make it better" I can't.

Revisions happen over email or WhatsApp — whichever you prefer. One distinction worth knowing: "raise the strings in the chorus" is a mix revision; "add a string line to the chorus" is an arrangement change — a different job I'd quote separately. The fixed quote you approve before I start covers the mix, the master and the revision rounds — no meter running while the notes go back and forth.

Louder used to win. It doesn't anymore. The big streaming services measure the average loudness of each song and turn everything to roughly the same level. A master crushed to sound loud gets turned down like everything else — and what's left is a song with no punch. The kick that was flattened to buy volume stays flattened; the advantage is gone.

Radio is a different pipe with the same conclusion: stations run their own heavy broadcast processing, so a master that's already slammed just gets squeezed twice. I master for impact at a sensible level — a song that hits hard because the mix breathes, not because the file is loud. In practice one good master serves streaming, radio and video.

The finished master as WAV and MP3 — the WAV for distribution, the MP3 for quick sharing. If you need stems of the final mix later, for live shows or a future remix, I provide them on request. All of it is covered by the fixed quote agreed before work started, so there are no surprises on delivery day.

Mixing and mastering is one part of what I do — I also write, compose, arrange and produce full songs, and my production services page explains how that works from the first melody to the final master.

If you want to know what your song actually needs, let me hear it. Send me your demo or rough mix at info@nicolasdebes.com or on WhatsApp at +961 71 809 950, and I'll tell you honestly whether it needs a mix, a master, or both.