📑 Sections (7)

1. The microphone and the signal

How to set up the NT2-A and the iD44: the mic's pattern, pad and filter, phantom power, gain, and how to set the level with the most intense part.

Vocals have no DI. On guitar and bass you can change the sound later without playing again; with vocals, what comes in through the mic is what you’ve got. The good news is that setting up the signal takes five minutes, and once it’s done you don’t touch it again.

The signal path
NT2-Acardioid
XLR cable
iD44 preampwith 48V
Mono trackfader at 0 dB

The NT2-A’s switches

The mic has three switches. They almost always stay in the same place.

Polar patterncardioid (omni and figure-8 aren't needed)

It defines which sides it picks up from. Cardioid takes mostly what's in front of it and rejects what's behind, so it leaves out a good part of the room. In an untreated room it's the only one that makes sense for vocals.

Pad0 dB (−5 or −10 dB if you scream hard)

It lowers the level before the mic's electronics. With normal vocals it stays at 0. Engage it only if, with the iD44 gain very low, the signal still gets close to clipping, which is possible with screams or growls at close range.

High-pass filter40 Hz (80 Hz if there's low-end noise)

It removes rumble the voice doesn't need: traffic, thumps on the floor, vibration from the mic stand. At 40 Hz you don't touch any of the voice. If the place has a lot of low-end noise, 80 Hz is still safe.

The iD44 and phantom power

The NT2-A is a condenser and needs phantom power (48V). There’s an order worth respecting:

  1. Lower the headphone and monitor volume.
  2. Connect the XLR cable to the mic and to the iD44 input.
  3. Turn on the 48V on that channel.
  4. To disconnect, the other way around: turn off the 48V, wait a few seconds and only then pull the cable.

That avoids the loud thump you can get from connecting with phantom on. The iD44’s 100 Hz filter and −10 dB pad stay off: you already have the mic’s, and it’s better to filter there.

The level

Peaks between −12 and −6 dBFS on the loudest phrase, same as the other instruments. What’s different on vocals is which part you measure with:

  • Sing the most intense part of the song, not the first verse. Many people set it with the quiet verse and the chorus clips on their best take.
  • Sing for real. People tend to sing louder in the real take than in the test.
  • When in doubt, stay close to −12. The voice is less predictable than a DI, and at 24 bits there’s plenty of margin.
🔴 Set this NOW· in recording

A vocal clip on your best phrase can't be fixed and forces you to sing it again, with a different energy. Lowering the gain a little costs nothing.

Studio One

  • Mono input with a clear name like “Vocal NT2-A”, in Song Setup → Audio I/O Setup.
  • 48 kHz and 24 bits, same as the rest of the song.
  • Fader at 0 dB. It doesn’t change what gets recorded, and this way the meter shows what’s coming in.

Check your take

Load a test take and look at the peak in the green zone and the spectrum. If in the silence between phrases you see a lot of low-end activity, it’s noise from the place: we’ll get to that in the next section.

Recording monitor

Play a signal or load your own DI/take and look at the level and the spectrum: that's how you know whether your take was recorded well before moving on to the mix.

This is how a well-recorded take looks and measures.
  • •Level — the peak should fall in the green zone (−12 to −6 dBFS): strong, with plenty of headroom.
  • •Clip — if it touches 0 dB (red) there's digital distortion: lower the interface gain and re-record.
  • •Spectrum — a metal rhythm guitar has body in the low mids and energy at 2–5 kHz; very high “fizz” gets tamed in the mix.
Load your own DI or take to check your recording against these references.
🎧 Audio to processChoose which sound goes through the effect:
  • •🎸 Bass — a synthesized riff with dynamics: ideal for hearing compression.
  • •📢 Pink noise — an even sound across the whole spectrum: ideal for “seeing” the EQ.
  • •🎵 Load your audio — your own file (MP3, WAV, OGG, FLAC, M4A…); it's processed locally in your browser and isn't uploaded anywhere.
Peak level (dBFS)— ·

🟩 Ideal zone: −12 to −6 dB. Avoid touching 0 dB (clip).

Raw vocal take for this sectionPending — this one is on you

I need a raw take of the verse vocal line of the guide song, unprocessed. Save it as prodmus_frontend/public/audio/voz/verso-voz-cruda.wav.

Format: WAV 24-bit / 48 kHz, mono, with peaks of −12 to −6 dBFS and no clip. It has to be your own recording, or one from someone who authorizes its use on the site.

Common mistakes

What NOT to chase yet

Quick path: signal ready

Quick path

  1. NT2-A on cardioid, pad at 0, filter at 40 Hz.
  2. Monitoring low, cable connected and only then 48V.
  3. iD44 filter and pad off.
  4. Mono input in Studio One, fader at 0 dB.
  5. Sing the most intense part and set peaks at −12 to −6 dBFS.
  6. Test take: listen to it and check it doesn't clip.

Expected result

Expected result

A clean signal with no clip and with margin, set with the most demanding part of the song. You shouldn’t need to touch the gain again for the whole session. What comes next is what weighs most on vocals: the room and where you stand.