Do vocal presets work, or does your voice need its own settings?
A vocal preset in Logic Pro is a set of frozen numbers tuned to someone else's voice, mic and room. Here is which settings depend on you, how to fit any preset in fifteen minutes, and when a profile measured from your own voice works better.
You load a vocal preset called something like "Radio Ready Lead" onto a track in Logic Pro, press record, read two lines, and play them back. The esses spit. In every pause a low hum you barely noticed on the dry track is now plain to hear, and the compressor meter hardly moves. Nothing in the chain is broken. The preset is doing exactly what it was built to do, for a voice, a microphone and a room that are not yours.
So, the short answer: vocal presets work as a starting point when your input looks like the input the preset was tuned on. The further your voice, mic, room and recording level drift from that, the more of the preset you have to re-tune, and past a certain point you are building your own settings with someone else's plug-ins already loaded. This article shows which numbers in a preset depend on you, how to fix them in about fifteen minutes in Logic Pro or GarageBand, and when a profile measured from your own voice beats any preset.
If you are not recording in a DAW at all and simply want your mic to sound better on calls, how to make your mic sound better without buying a new one is the page for you; it covers placement, gain and free fixes before any plug-in enters the picture.
What a vocal preset actually contains
In Logic Pro the thing most people call a vocal preset is a patch. Apple's guide describes a patch as the instrument, effects, Smart Controls and routing settings that control the sound of a track, and notes that patches for audio tracks can include default effects settings. So when you pick a vocal patch for an audio track, you are loading a stack of plug-ins with every knob already set: usually an EQ, a compressor, a de-esser, maybe saturation, and sends to a reverb or delay.
A plug-in preset is smaller. It sets one plug-in: one compressor, one EQ curve. A preset pack you download from a producer can be either kind. Many are Logic patches built only from stock plug-ins, so they load on any copy of Logic.
What neither kind contains is any information about the signal coming in. A preset is a set of fixed numbers. A threshold of -20 dB means the compressor starts working at -20 dB, whether your voice peaks at -6 dB or -30 dB. A boost at 4 kHz adds the same amount whether your mic is dark or already bright. The preset author chose those numbers while listening to their own source, and the numbers are frozen.
Where Logic Pro vocal presets live
People still search for "vocal presets Logic Pro X", although the Mac App Store listing today just says Logic Pro, and the current version is in the 12.x line. As of October 2026 it costs $199.99 and requires macOS 15.4 or later. GarageBand is free and uses the same idea of patches for audio tracks.
To find and audition vocal patches in Logic Pro:
- Select the audio track you record your voice on.
- Click the Library button in the control bar. If the track has no effects loaded yet, double-clicking the track header also opens it.
- Choose a category on the left and a patch on the right. The Up and Down Arrow keys step through patches, so you can play a loop of your recording and walk through them while it plays.
- Look at the channel strip in the left inspector. It shows the patch name and every effect plug-in the patch loaded. This is the part worth reading: it tells you what the patch will do to you.
Two Library features matter more for vocals than the patches themselves.
Patch merging. Choose Enable Patch Merging from the Action pop-up menu at the bottom of the Library. You get four buttons: MIDI effects, instruments, audio effects and sends. Deselect "audio effects" and you can try different reverb and delay sends from other patches while keeping the EQ and compressor you already tuned to your voice. Deselect "sends" instead and the reverse happens: your sends stay, and the effects chain comes from the new patch. This is how you stop a new patch from wiping out ten minutes of work.
User patches. Once a chain fits you, click Save in the lower-right corner of the Library and name it. Logic creates a User Patches folder, and that patch loads in any project. A saved patch of your own, tuned to your mic in your room, is the only vocal preset that was built for your input.
What a preset cannot know: your level, mic, room and voice
Every setting in a vocal chain makes an assumption about the input. Here are the usual ones, what they assume, and what you get when the assumption is wrong.
| Setting in the preset | What it silently assumes | What happens when it is wrong |
|---|---|---|
| Compressor threshold, e.g. -20 dB | Your voice peaks around a specific level | Record quieter and it never compresses; record hotter and it squashes every word and lifts the room noise in pauses |
| Noise gate threshold | A specific noise floor in a quiet room | Your fan or street sits above it and chatters through, or your soft word endings sit below it and get chopped |
| High-pass filter, e.g. 100 Hz | A voice of a certain depth | On a low voice it thins the chest; on a high voice it leaves room rumble untouched |
| Presence boost, e.g. +3 dB at 4 to 5 kHz | A darker microphone | A bright condenser gets harsher and the esses jump forward |
| De-esser frequency | Sibilance in one band | Your esses sit elsewhere, so it either misses them or dulls normal consonants |
| Reverb send | A dry, treated room | Your untreated room already adds reflections; the preset stacks a second room on top |
Only two rows, the high-pass filter and the de-esser, are about the voice itself. The other four are about level, mic and room, and those change the moment you move the mic 10 cm or record at night with the heating on.
Level is the hidden variable
The most common reason a preset "doesn't work" has nothing to do with taste. It is input gain. Compressors, gates, saturation and de-essers all react to level. The preset author recorded at some level; if you are 12 dB quieter, the compressor is effectively switched off and the de-esser rarely triggers. If you are 6 dB hotter, everything works too hard.
So before judging any preset, set your gain so your loudest words peak somewhere around -12 to -6 dB on the track meter. That one step makes most presets behave closer to what their author heard. Our article on which microphone settings actually change how you sound goes through gain in more detail.
Fitting any preset to your voice in fifteen minutes
This works in Logic Pro and, with fewer controls, in GarageBand. You need headphones, a quiet five minutes and a test recording.
Record a test take first. Read about a minute of text out loud, at your normal distance from the mic. Include plenty of S, SH and T sounds, a few P and B words close together, a soft passage and a loud one. Then sit in silence for five seconds before stopping. Loop this take while you adjust. You are tuning to a fixed recording, which is far easier than tuning while you talk.
Then go through the chain in this order:
Step Plug-in What to set How you know it is right
1 Input gain Loudest words peak about -12 to -6 dB Meter never touches 0 dB
2 High-pass Raise until the voice starts to thin, Rumble in the silent 5 s drops,
then back off a little voice keeps its weight
3 Noise gate Threshold just above the noise in Room noise gone in pauses,
(or remove) the silent 5 s; skip it if possible word endings intact
4 EQ Cut before you boost; sweep for a Fewer dB moved than the preset
boxy or ringing spot and dip 2-4 dB had, voice still sounds like you
5 Compressor Lower threshold until loud words get Gain reduction meter moves on
about 3 to 6 dB of gain reduction phrases, rests in pauses
6 De-esser Find your sibilance band, then set Esses calmer, T and K still crisp
strength by ear
7 Reverb/delay Lower the send until you only notice Words stay intelligible
it when you mute it
A few notes on the steps that go wrong most often.
The gate. On a recording you can edit, a gate often does more harm than good, because a tight threshold eats breaths and the tails of words. If the room noise in the silent seconds is low after the high-pass filter, delete the gate. Our guide to noise gate settings in OBS explains the threshold and hold controls if you are using one live.
The de-esser. Logic's DeEsser lets you set the frequency it listens to (the Detector) separately from the frequency it turns down (the Suppressor). Set the Monitor pop-up to Det and you hear only what the detector hears; sweep the Detector frequency until that solo is mostly your esses. Then switch Monitor back to Off and raise Strength until the esses stop jumping out. Apple describes DeEsser as compressing the band dynamically rather than cutting it statically, so the voice does not get darker when there is no sibilance.
The EQ. If you find yourself boosting more than a few dB anywhere, check the mic position before you check the EQ. A boost often patches a problem that sitting 10 cm closer, or turning the mic slightly off-axis, would remove at the source.
When the chain fits, save it as a user patch with a name that records the conditions, for example "Vocal, USB mic, bedroom, 15 cm". Change the mic or the room and you will want a new one.
Singing presets and speaking presets are different jobs
A lot of vocal presets for Logic are built for a sung lead in a full mix. They carry heavier compression, a bright top end so the voice cuts through instruments, and long reverb or delay sends. Put the same chain on a spoken voice with no music under it and every one of those choices is exaggerated: the brightness has nothing to compete with, the compression lifts every breath, and the reverb makes a podcast sound like it was recorded in a stairwell.
The reverse also holds. A spoken-word chain is gentler and drier, and on a sung chorus it may leave the voice sitting behind the backing track.
So filter presets by job before you filter by name. If you are recording songs at home and still deciding on the hardware, which microphone is best for recording vocals at home covers that half of the problem; a preset cannot make up for a mic that picks up more room than voice.
The catch most people hit: presets stay inside the DAW
A Logic patch processes audio inside Logic. Zoom, Telegram, Discord, Google Meet and OBS do not see it. They take the raw signal from your microphone. Tune a chain you love in Logic, join a call, and the other side still hears the dry mic.
There are three ways around it on a Mac, each with a cost:
- Route Logic's output into a loopback driver. An open-source driver such as BlackHole creates a virtual audio device that one app can play into and another can use as its input. You keep Logic open with input monitoring on during every call, and you manage the routing yourself. Our comparison of Audio Hijack and Loopback covers the paid tools built for this.
- Rebuild the chain in OBS if you only need it for streams and recordings made in OBS. OBS has its own per-source filters, with no Logic involved. The order matters there too: see which OBS mic filters to use, and in what order.
- Use a virtual microphone that does the processing itself, which is the approach covered in the next section.
Presets versus settings measured from your voice
There is a middle category between a frozen preset and tuning everything by hand: tools that listen to your actual signal and build the starting point from it.
In the mixing world, iZotope's Nectar 4 has a Vocal Assistant that, in iZotope's words, creates custom presets to fit your specific track. It runs as a plug-in in your DAW, so it is for recorded vocals you are mixing, and its output still lives inside Logic.
For your live voice in every app, TunedMic moves the same idea out of the DAW entirely. It adds a virtual microphone called TunedMic to the Mac. Setup is a Sound Check: you read a short passage aloud for about a minute, and while you read, TunedMic takes more than 30 measurements of your voice, your mic and your room. From that it builds nine profiles in three families, each one calculated from your Sound Check rather than loaded as a stock preset:
- Narrator, for calls and lessons: Natural, Warm, Focused
- Creator, for streams, videos and podcasts: Balanced, Broadcast, Crisp
- Vocal, for singing and practice: Gentle, Open, Forward
Any app that lets you choose a microphone can pick TunedMic, so the processed voice reaches Zoom, Telegram, Discord, OBS and the rest at once, and switching profile changes the sound everywhere with nothing to restart. If you want to adjust by hand there are optional controls for noise, room echo, clarity, loudness and a 12-band equalizer; you never have to touch them. It adds less than a tenth of a second of delay, it runs locally, and your voice is never uploaded.
Here is how the three approaches compare:
| Logic patch or downloaded preset | Analysis plug-in (e.g. Nectar 4 Vocal Assistant) | TunedMic profile | |
|---|---|---|---|
| Built from | The author's voice, mic and room | Your recorded track | A one-minute Sound Check of your voice, mic and room |
| Where it runs | Inside the DAW | Inside the DAW | System-wide, as a virtual mic |
| Reaches Zoom, Telegram, Discord | Only with loopback routing | Only with loopback routing | Yes, pick TunedMic as the mic |
| Best for | Mixing songs and voice-overs you edit | Mixing recorded vocals | Live voice on calls, streams, recordings and singing practice |
| Platform | Mac (Logic, GarageBand) | Whatever DAWs the vendor supports | Mac with Apple silicon, macOS 14 or later |
The limits matter. TunedMic needs a Mac with Apple silicon (M1 or newer) and macOS 14 or later; there is no Windows version and Intel Macs are not supported. It is not a mixing tool: if you are producing a song in Logic and want the lead vocal to sit in a dense mix, a tuned channel strip or a plug-in like Nectar is the right instrument. And the measurements describe the mic and room you had while reading, so after a mic change it makes sense to run Sound Check again, the same way you would save a new user patch.
Before you buy another preset pack
Another preset pack is always one download away, and sometimes one of its presets lands close to your voice on the first try. When it does, the reason is usually that your mic, room and level happen to resemble the author's. When none of them land, buying a bigger pack will not change that; the fix is the fifteen-minute pass above, done once per setup and saved.
A simple test tells you which situation you are in. Load the preset, set your gain properly, and watch the compressor's gain reduction meter while your test take plays. If it moves 3 to 6 dB on loud phrases and rests in the pauses, the preset is in the right neighborhood and only needs EQ and de-esser tweaks. If it sits still or never lets go, the preset was built for a different level and every setting after it is reacting to the wrong signal.
Your next step
If your goal is recorded vocals in Logic Pro or GarageBand: record the one-minute test take today, run it through the seven steps in the chain above, and save the result as your own user patch with the mic and room in its name.
If your goal is your live voice on calls, streams or singing practice: watch the 38-second video on the TunedMic home page, which covers the one-minute setup and the Speech Report after a call. Then run Sound Check on your own Mac during the free 7-day trial, which starts at first launch, includes every feature and asks for no card or account. After the trial a license is $19 one-time; the details are on the pricing page.