How to Use Beat Stems

When you buy a beat from me, you get a zip file with several pieces inside. The stems are the main event. The other files exist to help your engineer mix the song properly and to give you something to vibe to before you record. Here’s a walkthrough of what’s in the zip and how to use each piece.

What’s in the Zip

You’ll find four things when you unzip the file:

• A reference stereo mix that isn’t mastered
• A lightly mastered version
• A text file with notes for your engineer
• The stems (zip file)

I’ll go through each one below.

The Reference Stereo Mix

This is the mixed version of the beat with no mastering applied. Your engineer uses this to hear what I heard when I made the beat. Think of it as a target your engineer can aim for or improve on. (Technically you could record vocals over this version, but you’d probably have to use EQ to ensure there is a pocket carved out for your vocals.)

The Lightly Mastered Version

I push the volume up on this one so it hits hard in the car or in your headphones. It’s for vibing and getting yourself in the right headspace before a session. Don’t record over this one. The stems are where the actual mix happens.

The Notes File

The text file gives your engineer the technical details they need to set up the session fast. It has the track name, the current song arrangement, the BPM, and the key. Forward this file to your engineer before your session starts so they can drop everything into place before you even arrive. My email address is in there too so they can reach out with any questions. 

The Stems

All stems export at 24-bit, 48 kHz. In my opinion, that’s the floor for getting a professional result. Anything below that and you’re starting the mix with a handicap. I could export them at 96 kHz if you need it, but unless the source material is at 96 KHz, not sure it is required. (If you are recording vocals at 96 kHz, then your engineer should be able to easily upsample them.)

Each stem represents one sound source. I split out every drum element separately and every musical instrument separately. So instead of one combined drum track, you get the kick, the snare, the hats, the 808, and so on. Same with the instruments. That gives your engineer full control over every piece of the beat.

How I Name the Stems

I prefix every file so your engineer knows what they’re looking at without opening a single one.

Drum files start with “Drums.” Instrument files get prefixed with the source I used. If the part came from Kontakt, the file starts with “Kontakt.” Same for Serum, SubXL, or any other virtual instrument. If I used a sample, the file gets tagged with where the sample came from. Splice samples get a “Splice” prefix. Samples pulled from Maschine get a “Maschine” prefix.

All samples I use are royalty free, so clearance is never an issue. I’m not trying to get sued, and neither are you. The contract covers this too. Even in the unlikely case I got something wrong on a sample, you’re indemnified (aka not held liable).

Here’s a real file list from one of my beats. Fifteen stems total, including all the drum samples, the Kontakt and Serum2 instruments, and one Maschine sample:

• Drums – 808 (Drench) (C).wav
• Drums – AWB2 HH 7.wav
• Drums – AWB2 HH 10.wav
• Drums – AWB2 OHH 2.wav
• Drums – AWB2 OHH 4.wav
• Drums – AWB2 Snare 7.wav
• Drums – AWB2 Snare 15.wav
• Drums – BricksDaMane Open HiHat 8.wav
• Drums – Kick (HeftyBoy).wav
• Drums – Snare Guccy 3.wav
• Kontakt_Feelit_Pianohaus.wav
• Kontakt_Homage_ChoppedKeys.wav
• Maschine_Sample.wav
• Serum2_KeysCerealBoxPrize2023.wav
• Serum2_Module2K_OlDirtyGrand.wav

If you ever need the exact sample name from one of my tracks, just ask. I can pull that information for you. I don’t always rename every track before export, but the information exists.

Why I Export Stems This Way

When I track out a beat, I reset every fader to default before I bounce. That gives your engineer the strongest signal possible to work with. I also strip filtering and effects off the individual tracks. The only time I leave processing on a stem is when that processing is the sound of the track. If a filter sweep is the vibe of the song, it stays. Otherwise it comes off.

Think of mixing like sculpting ice. You want the biggest block you can get. Your engineer can shape and refine from there. Hand them a small chunk and they have nothing to work with.

I also avoid pre-made loops when I can. Building the parts myself from instruments and samples gives me full control over how every element moves through the beat. A loop locks you into someone else’s decisions about rhythm, dynamics, and arrangement. When I play the part in, I can shape it around the rest of the track instead of working around the loop.

I also export every stem from the start of the session. So your engineer drops the files into their DAW, left-aligns everything, hits play, and the song plays back in sync. No lining up, no guesswork.

Why It Sounds Bad on First Playback

When your engineer first hits play on the raw stems, the beat is going to sound rough. Drums too loud. Keys panned dead center. The 808 overpowering the whole track. That’s by design. I reset all the volume and pan settings before export so your engineer can build the mix from scratch around your vocals.

The reference mix is there for exactly this reason. Your engineer can listen to my mix, hear what I was going for, and use it as a starting point. A good engineer often beats my reference because they have more time on one song than I spent and they’re mixing to your voice and your performance.

That’s the actual difference between sounding like a record and sounding like someone rapping over a YouTube beat. When you pull a beat off YouTube or a low-tier lease site, you’re stuck with whatever the producer printed as a stereo file. You can run that file through stem separation tools like iZotope Ozone or LALAL.ai and pull out the snare or bass track, but the quality isn’t the same. Stem separation works by using AI to identify each sound in the mix, but artifacts will show up the second your engineer tries to do real work on those tracks. Boost the snare and you’ll hear smearing. Cut the bass and the kick takes a hit with it. You’re working with a reconstruction, not the source. Not only that, you are probably using an MP3 file which already has its own degradation. Any, I digress.

Original stems are the source. Every sound is exactly what came out of my session, untouched by any algorithm trying to figure out what it used to be. Your engineer can properly boost, cut, side-chain, filter, or rebalance anything without fighting artifacts. That’s why the same engineer can make a song from my stems sound better than a song from a separated YouTube beat, even if they’re equally skilled in both cases. The ceiling is just higher when the floor is real source material.

How to Test Your Mixing Engineer

Here’s a simple way to gut-check your engineer’s work. After they finish the mix, ask them to bounce the beat alone into a stereo file without your vocals. Compare that bounce to the reference mix I sent in the zip.

The engineer’s version should sound at least as good as my reference. Ideally better, because they’re working with the same source material I had plus more focused time on one song. If their version sounds worse than mine, that tells you something.

Mixing is craft work built on years of ear training and decision-making. Not every engineer is good at it. The wrong engineer will make even the cleanest stems sound mid. Hopefully your engineer passes the test and your song sounds even better than what I sent.

Final Word

Everything in that zip exists for one reason. I want your record to win. Use the reference for direction. Use the lightly mastered version to vibe. Use the notes file to brief your engineer. And put the stems in the hands of someone who can mix.

I’ve worked this way for years. Engineers tell me my stems save them hours of prep, which means more of their time goes into mixing your song. That’s how a beat becomes a record.