Vocal Stems Explained: Dry, Wet, Lead, Doubles, Harmonies (and How Producers Use Them)

A vocal stem is one part of a vocal production exported as its own audio file, aligned to the song's timeline, so you can mix it independently of the other parts. Where an acapella gives you the whole vocal as one file, stems give you the lead, the doubles, the harmonies and the ad libs as separate tracks. This guide explains every file you will find in a professional vocal package, what each one is for, and how producers and DJs actually work with them.

What a vocal stem is (and what an acapella isn't)

Producers use "stem" loosely, so here is the precise version. A stem is a submix of related tracks bounced to one file. In a full song, the drum stem contains all the drums, the vocal stem contains all the vocals. Inside a vocal package, the word goes one level deeper: the lead vocal is one stem, the doubles are another, the harmonies another. Every stem starts at the same point and runs the same length, so when you import them all they line up without any editing.

An acapella is the finished vocal mix as a single file: lead, doubles, harmonies and effects already balanced. It is faster to work with and it is exactly what you want when the vocal already sounds right and you are building around it. It is also all you get. You cannot turn the harmonies down, mute a double in the verse or pitch the ad libs without affecting everything else.

Stems are the flexible version, acapellas the convenient one. A good package gives you both, which is what the rest of this article is about.

What's in a professional vocal package

On Voxalized every purchase includes at minimum the lead vocal dry and wet as lossless WAV files, with your personalized license agreement as a separate PDF download next to them. The Professional license adds everything else the singer recorded: the premixed vocal, doubles, harmonies, backings, production stems, MIDI and lyrics. Here is what a Professional download looks like when you unzip it.

Song Title – Professional.zip
├─ README.txt              title, singer, BPM, key, genre, license tier
├─ Lead vocal (dry)/       the tuned, edited lead without effects
├─ Lead vocal (wet)/       the same performance through the singer's chain
├─ Premixed vocal (wet)/   all layers balanced into one file
├─ Doubles/
├─ Harmonies/
├─ Backings/
├─ Production stems/       ad libs, FX and any extra layers
├─ MIDI/                   what the singer exported from the session
└─ Lyrics file/            TXT or PDF

Song Title – License Agreement.pdf   (its own download, next to the ZIP)

Inside the folders the files keep the names the singer exported them with, so you see their labeling rather than a generic one. Key and BPM are in the README and on the product page, so you know before you import whether the vocal fits your track. Which folders show up depends on what the singer recorded: how many files land in Doubles, Harmonies or Production stems changes from vocal to vocal, which brings us to the parts.

Lead dry vs lead wet

The lead dry is the tuned, edited, cleaned performance with no reverb, no delay, no creative processing. The lead wet is the same performance run through the singer's own vocal chain.

We ship both for a reason. The singers on this platform are professionals; several of them have releases with hundreds of millions of streams, and their vocal chains reflect that. If you are starting out, the wet lead is a finished, radio-ready vocal you can drop in and build around today. If you have your own chain and your own taste in reverb, the dry lead gives you a blank canvas: your delay, your space, your saturation, nothing baked in that fights your mix. Most experienced producers start dry and keep the wet version open on a muted track as a reference for how the singer heard it.

Doubles, harmonies, backings, ad libs

Doubles are the lead sung again, usually once or twice, panned or tucked underneath to thicken the chorus. Harmonies are different notes sung against the lead, thirds and fifths mostly. Backings are the "oohs," "ahs" and response lines that fill the space between lead phrases. Ad libs are the improvised runs and shouts, often at the end of a section.

How many of each you get depends on how the singer built the song. Some vocals come with four backing layers, some with one. That is not a quality difference, it is an arrangement decision, and the product page tells you what the package includes before you buy.

Premixed, MIDI, lyrics: why they're in the folder

The premixed file is all the stems balanced together, essentially a wet acapella. It is your fastest path to a demo.

The MIDI folder holds what the singer exported from the session: usually the chords or the piano part the vocal was written over, sometimes a bassline or the vocal melody itself. It carries little weight in the final track and a lot of weight in the first ten minutes, because it tells you instantly what the harmony is doing.

The lyrics file is for your metadata, your lyric video and your own sanity when you are deciding where the chorus starts.

Five things you can only do with stems

Remix the arrangement. Mute the doubles in the verse so the chorus opens up, or drop the harmonies for the first half and bring them in on the second.

Process each part differently. A tight, dry lead with a huge, washed-out harmony bed is a production choice you cannot make with one acapella file.

Chop and re-pitch without collateral damage. A dry lead slices cleanly. A mixed acapella brings reverb tails and backing notes into every slice. We covered this in vocal chops.

Change key or tempo cleanly. Pitch-shifting or stretching a single dry voice sounds far better than shifting a stack of layered voices with effects.

Sidechain and automate per part. Duck the backings under the lead, ride the ad libs up at the drop, and leave the lead untouched.

How to work with stems in your DAW

Import, align, check the tempo

Set your project tempo to the vocal's BPM first, then drag every stem in at bar one. If the package is built correctly, nothing needs nudging. Check that the wet and dry lead line up sample-accurately; if you plan to blend them, any offset will phase.

Chords and bass from the MIDI, then structure, then drums

Here is the order I use. Vocals and MIDI go in together. I open the MIDI on a piano and play through it to hear where the chords sit and what the bassline is doing. From there the structure comes quickly, because the vocal already tells you where the verses and choruses are: I block out the sections, drop in drums to get the groove under the vocal, and I have a rough idea of the whole track within an hour. Having key, BPM and MIDI up front is what makes that possible. Speed matters in this business. When a new cover vocal goes live, the producer who gets a demo out in a day has a real advantage over the one who spends a week figuring out the chords.

Start dry, use wet as the reference

Mute the wet lead, work on the dry one, and A/B against the wet version now and then. If your processed dry lead sounds worse than the singer's wet lead, you have learned something about your chain. If it sounds better, you have a vocal that is truly yours. Either way the wet version is your safety net for the demo.

Gain-stage before you touch a plugin

Stems from a professional session are usually exported hot. Pull each stem down so the lead peaks around -6 dBFS and the backings sit well under it before you add anything. Half of "my vocals sound harsh" problems are stems hitting a compressor too hard. From there, the chain in our guide to mixing vocals applies to each stem separately.

Stems for DJs

DJ software has caught up with the word. Serato DJ Pro's Stems, Rekordbox and djay all separate a track into vocals, melody, bass and drums in real time, and for live mashups that is impressive technology. It is also a machine-learning approximation of a vocal, with artifacts that show up the moment you solo it in a quiet intro.

A real vocal stem, exported from the session, has none of that. It is the actual recording. For DJs who produce their own edits and tools, a licensed dry lead is the difference between a mashup that sounds like a mashup and one that sounds like a record.

The license also covers what a separated stem does not: the right to release the result.

Stems and licensing (can I share them?)

Buying the Professional tier gives you the stems for use in your own productions. It does not let you pass them on. Sending a dry lead to your mixing engineer for your track is normal use; uploading the stems to a remix contest, a sample site or a group chat is redistribution and is excluded. The finished track can go anywhere. The stems stay with you. The full picture of what the license does and does not cover is in royalty-free vocals explained.

The questions we actually get

Producers rarely ask about stems themselves; the files are self-explanatory once you open the folder. The two questions that do come up are about stem counts and about covers.

On stem counts: a vocal with one backing layer is not a lesser product than one with four. It is a different song. Check the product page for what is included.

On covers: when the vocal is a cover, the stems are the singer's performance and the composition still belongs to the original writers. You release it like any other cover, and our guide on how to legally release a cover song walks through the distributor steps.

If you want the maximum control over how the vocal sits in your track, the Professional tier is the one to get, and the FAQ covers the edge cases on downloads and re-downloads.