How to bounce acapella, instrumental and full-mix versions from one Logic Pro session

You can bounce an acapella, an instrumental and the full mix from one Logic Pro session without rebuilding anything, as long as the mixer is organised into two summing stacks: Vocals and Instruments. The instrumental is the Vocals stack muted. The acapella is the Vocals stack soloed. The full mix is nothing touched. The same two stacks give you vocal-up and TV mix variants, and they are the structure your stems come from, so every version lines up sample for sample.

I set up every vocal session this way before a library delivery. Here is the routing, the trap that catches almost everyone, and the bounce settings.

The two-stack setup

Select every vocal track: lead, doubles, harmonies, ad libs. Choose Track > Create Track Stack (Shift-Command-D) and pick Summing Stack. Logic routes them all to a new aux with one fader, mute and solo. Name it Vocals.

Do the same for everything else and name it Instruments. If you also deliver stems, make sub-stacks inside Instruments for Drums, Bass, Guitars, Keys, Synths and FX, so the instrumental and the stems come from the same tree.

The whole song is now two faders at the top level. Mute Vocals and you have the instrumental. Solo Vocals and you have the acapella. Any processing on the Vocals aux stays in both versions, so the balance you mixed is the balance you deliver.

If the library wants a TV mix, the full mix with the lead vocal removed but backing vocals kept, go one level deeper. Inside Vocals, create a Lead sub-stack and a BVs sub-stack. The TV mix is Lead muted, everything else playing.

The send-return trap

This is where most instrumental bounces go wrong.

Say your lead vocal sends to a reverb on an aux called Vox Verb, and that aux lives outside the Vocals stack. Mute the Vocals stack and, with post-fader sends, the send goes silent too, so the reverb stops. But if any send is pre-fader, the reverb aux keeps receiving the vocal while the stack is muted, and your instrumental has a ghostly wet vocal in it. The reverse happens if an instrument shares the reverb aux and you solo Vocals for the acapella: in my experience Logic keeps the shared return audible, so the acapella carries a trace of the piano's reverb.

The fix is routing, not remembering. Move the vocal reverb and delay auxes inside the Vocals stack so they mute and solo with it, or give the vocals dedicated returns that nothing else feeds, with post-fader sends. Then listen to the end of the chorus in the instrumental and confirm there is no tail.

The post on exporting stems from Logic Pro goes further into where sends end up when you export rather than bounce.

The versions and how each one is set

With the stacks in place, each version is a mute or solo state. Write them down so they are repeatable.

  • Full Mix. Vocals stack: Playing. Instruments stack: Playing. Notes: Nothing changed.

  • Instrumental. Vocals stack: Muted. Instruments stack: Playing. Notes: Check for vocal reverb tails.

  • Acapella. Vocals stack: Soloed. Instruments stack: Muted. Notes: Keep the Vocals aux processing in.

  • Vocal Up. Vocals stack: +1.5 to +2 dB. Instruments stack: Playing. Notes: Lead vocal louder for dialogue-free spots.

  • TV Mix. Vocals stack: Lead sub-stack muted. Instruments stack: Playing. Notes: BVs stay in.

Some editors want a vocal-up mix so the vocal survives being ducked under a voiceover. A 1.5 to 2 dB lift on the Vocals aux fader is usually enough. Do it on the stack fader, not the individual tracks, so the internal blend stays the same.

For unmastered versions of any of these, bypass the limiter on the Stereo Out before you bounce. That routine is in how to bounce an instrumental and an unmastered mix.

Bouncing each version

Every version uses identical bounce settings so the files match in length and line up.

  1. Set the cycle range from bar 1 to the project end, past the last tail.

  2. Set the mute and solo state for the version from the table above.

  3. Choose File > Bounce > Project or Section (Command-B).

  4. Destination PCM, File Format WAV, Resolution 24-bit, Sample Rate at the project rate (usually 48 kHz for library work). Normalize Off. Mode Offline. Include Audio Tail on.

  5. Name the file to your convention and bounce.

  6. Change the mute and solo state for the next version and repeat.

Do not change anything else between bounces. Not a fader, not a plug-in, not the locators. If you need a tweak, make it and re-bounce every version.

For the acapella, soloing Vocals also solos the auxes it feeds, so the vocal reverb stays in. That is what you want for a library acapella. If a remixer asks for a dry acapella, bypass the vocal sends and bounce again under a different name.

Checking each bounce

Length. All versions should be the same duration to the sample. If the acapella is shorter, a tail got cut or a locator moved.

Content. Play the first and last ten seconds of each. The instrumental should have no vocal reverb tail. The acapella should have no bleed from a shared return. The TV mix should have backing vocals in the chorus.

Null, roughly. Import the full mix and the instrumental into a fresh project, flip polarity on one with Gain, and the result should sound like a slightly odd acapella. With a limiter on the master it will not be exact. You are listening for drums leaking through.

Naming

Use the library's pattern and keep the version word consistent:

Midnight_Run_Full_Mix.wav
Midnight_Run_Instrumental.wav
Midnight_Run_Acapella.wav
Midnight_Run_Vocal_Up.wav
Midnight_Run_TV_Mix.wav

Decide once whether it is "Acapella" or "Vocals Only" and never vary it. The stem delivery checklist has more on conventions and why ingest systems care.

Where QuickStem fits

The setup above is what QuickStem looks for. It scans the Logic session's mixer and routing to find the stem structure, including summing stacks like Vocals and Instruments. I review what it found, keep the groups or combine channels into custom stems, and pick the mix versions: Full Mix, Instrumental, Unmastered Full Mix, Unmastered Instrumental and custom mixes. Acapella and TV mix are custom mixes I define once and save in a template with my naming convention. Vocal-up needs a fader move, so that one I still bounce by hand.

Bounce All then runs every version in one go. QuickStem drives Logic through Apple's Accessibility API, setting the mutes and solos and bouncing each version the way I would by hand. It never touches audio files directly, and it verifies the session is unchanged when it finishes. Files land in STEMS, MIX DELIVERABLES and DOCUMENTS folders, named to the pattern I set. Several tracks go in the batch queue and run overnight.

It does not fix bad routing. If a pre-fader send leaks vocal reverb into your instrumental by hand, it will leak in QuickStem too.

QuickStem runs on macOS 14 or later with Logic Pro 11 or later (verified on Logic Pro 12). It is signed and notarized, and one licence covers two Macs. Details are on the QuickStem product page. You can buy QuickStem for $49 or QuickStem Plus for $79 before the opening sale ends November 30, 2026. Every purchase includes one free Soundwall master. Ableton Live and Pro Tools versions are coming.

FAQ

How do I bounce an instrumental in Logic Pro?
Put every vocal track in a summing stack, mute the stack, and choose File > Bounce > Project or Section with the same locators and settings as the full mix. Check the end of the chorus for vocal reverb tails.

How do I bounce an acapella in Logic Pro?
Solo the Vocals summing stack and bounce with the same settings. Soloing the stack keeps the vocal reverb and delay returns audible.

Why does my instrumental still have vocal reverb in it?
A pre-fader send, or a reverb aux shared with other instruments, keeps feeding the return while the vocals are muted. Move the vocal returns inside the Vocals stack or give them dedicated auxes with post-fader sends.

Can QuickStem bounce acapella and instrumental versions together?
Yes. Instrumental is a built-in version, and acapella and TV mix are custom mixes you define once. Bounce All renders them in a single run with your naming convention. Vocal-up needs a fader change, so bounce that one by hand.

Author: John Coggins is the founder of Soundwall Studios in Westlake Village, CA, and the creator of QuickStem. He's a songwriter and producer with over 1,000 works registered with ASCAP and a long-time writer for Extreme Music.

Previous
Previous

Bounce in Place vs Export Tracks as Audio Files in Logic Pro: which one for stems?

Next
Next

How to make 60, 30 and 15 second edits of a library track in Logic Pro