The Vocal Chain That Makes a Demo Sound Like a Record

The single biggest reason a home-studio demo sounds like a demo is not the room. It is not the microphone. It is the vocal chain — the specific order, settings, and small judgment calls between the singer and the bounce. Most home recordings get the room about as right as they can afford to. They get the mic close to a reasonable industry standard. And then the signal chain falls apart, usually in the same five places.

What follows is the working vocal chain I use on a typical Mollohan Production session, lightly adapted for a home setup. Nothing here is exotic. The point is the order and the gain staging, not the gear list. A reader who follows this chain with a $200 USB interface and a $99 condenser will end up with a more usable vocal than one who follows nothing at all with a $3,000 mic.

Step 1: Pick the right mic for the voice, not the room

The mic-to-voice match is the first decision and the one most home recordists get wrong. The default move — reach for the large-diaphragm condenser because it's the most expensive thing on the stand — is fine for about half of singers and wrong for the rest.

• Bright, breathy, mid-rangey voice: a large-diaphragm condenser is usually right. A neutral one — not one that's already aggressively pre-EQ'd for sibilance.
• Loud, hard, fundamentals-heavy voice: a dynamic mic like the Shure SM7B or a Sennheiser MD 421 will almost always sound more like a record. The condenser will exaggerate everything the voice is already doing.
• Quiet, intimate, close-mic'd folk vocal: a ribbon will sometimes get you there in a way a condenser can't. They're harder to track at home but worth experimenting with.

The wrong fix most people try first: spending more money on a mic before doing this audit. A $300 condenser correctly matched to the voice will out-record a $2,000 condenser that's wrong for the voice. Every single time.

Step 2: Set the preamp gain so you have headroom for the chorus

The number one home-studio mistake: setting input gain on a quiet verse and getting clipped on the chorus. The fix is simple. Track the loudest part of the song first. Set gain so peaks live around -10 dBFS on that part. Everything else will sit comfortably below.

If the interface preamp is noisy when the verse is quiet, that is not a gain-staging problem; it is an interface problem. Do not solve it by pushing more gain into the chorus. Solve it by getting closer to the mic on the verses, or by using a cleaner preamp on the way in. The vocal that holds up at mix is the one that was tracked with headroom, not the one that was tracked hot to "make it sound bigger".

Step 3: Print a small compression on the way in

The home-studio orthodoxy of "track flat, compress later" is almost right, but in practice it costs the recording its character. A small amount of compression on the way in — three to five decibels of gain reduction on the loudest moments, slow attack, medium release — does two things. It catches the peaks so you can ride the fader less in the mix. And it lets the singer hear themselves more like a record while they're tracking, which changes how they sing.

A clean optical or FET compressor is what you want here. Not a glue bus compressor. Not a vintage tube emulation set hard. Just a few dB of correction that the singer can hear in the headphones. If a plugin is the only option, an LA-2A emulation set conservatively is fine. The point is the singing, not the compressor.

Step 4: Subtractive EQ before additive

The biggest EQ mistake I see in home-studio mixes is the high-end boost added before the low-mid problem has been removed. Subtract first.

1. High-pass around 80–100 Hz. Nothing useful below that on a vocal.
2. Find the boxy frequency — usually somewhere between 250 and 500 Hz — by sweeping a narrow boost until something sounds bad, then cutting at that frequency by 2–4 dB.
3. Find the harshness — usually 2–4 kHz — and treat it the same way. A narrow 1–2 dB cut is almost always enough.
4. Only then consider whether the voice needs anything added on top. Most don't.

The vocal that ends up sounding "expensive" almost always has more cuts than boosts. Beginning mixers reach for the high shelf first; experienced ones reach for the parametric.

Step 5: De-ess after, not during

Tracking with a de-esser on costs you flexibility you'll wish you had. Sibilance is easier to fix in the mix than to recover after it's been over-processed on the way in. Track flat for sibilance. In the mix, use a dynamic EQ or a narrow-band de-esser on the offending frequency (usually 6–9 kHz on a female vocal, 5–7 kHz on a male vocal), and ride it where it matters most.

If this happens…

• The vocal still sounds thin after all of this: the room is fighting you. Move the mic and singer further from the back wall, hang a blanket behind the singer's head, and re-track. A thin vocal is almost always a reflection problem, not an EQ problem.
• The vocal sounds nasal: probably a 700 Hz buildup. Try a narrow cut there, 2–3 dB.
• The vocal pumps under the chorus drums: the compressor's release is too fast. Slow it down until the breathing stops.
• The singer can't find the note: turn off the reverb in the headphone mix. Add it back at the end.

Once you've got it working

Save the chain as a template. Every Mollohan Production session opens with the same template — the same plugin order, the same starting points, the same headroom. The settings change on every record, but the chain doesn't. That consistency is what lets us audition fifteen different artistic choices in a session without losing track of which one is working.

If you're a writer on the MPIArtist development side reading this, the practical version of this is simpler: ask your producer to print you a rough mix with the vocal chain locked, and use that as your reference for the rest of the song. The instinct to chase a "better" vocal sound on every demo costs records more than it saves them.

Frequently asked

What's the cheapest version of this chain that still works?

A $99 large-diaphragm condenser, a clean USB interface with a usable preamp (Focusrite Scarlett-class is fine), and stock DAW plugins. Plugin tier does not matter as much as plugin order. The chain above will work with stock compressors and EQs.

Do I need to record with compression on, or can I add it after?

Either works. Print a little compression on the way in if you want the singer to hear themselves more like a record. Track flat if you'd rather make every decision in the mix. The result can be identical; the singing usually isn't.

How loud should I track the vocal?

Set gain so the loudest moment in the song peaks around -10 dBFS. That gives you headroom for the chorus and keeps the noise floor well below the singer.

Is the SM7B really worth it for home recording?

For a loud or fundamentals-heavy voice, yes. For a quiet folk vocal, often no — a condenser will get there faster. The mic-to-voice match matters more than the brand.

How does this differ from the chain you'd use in a treated studio?

The chain itself doesn't change much. What changes is how much subtractive EQ the vocal needs — a treated room delivers a cleaner signal, so the early cuts are smaller. The order and the gain staging are the same.