When a vocal and a guitar share one track, you mix them as a single instrument: every EQ, compression and effects decision is a compromise serving both sources at once. You cannot separate them, so you stop trying, and you aim the tools at specific problems (a de-esser at string zing, a filter at foot stomps) rather than at "the vocal" or "the guitar."
On Dr. Dog's "Fat Dog," Scott McMicken recorded his lead vocal and acoustic guitar to the same track of an 8-track tape machine. Matt Ross-Spang's reaction is the first lesson: this would annoy a lot of engineers, but the artist made a decision, committed to it, and loved the result. As Matt puts it, if the artist loves what they brought me, I love it too.
The chain, problem by problem
Every processor on this track targets something specific:
- Surgical EQ first: narrow cuts to tame foot stomps captured in the recording, without removing them entirely, because the stomps are part of the vibe. Gain lowered 7 dB for headroom.
- A de-esser aimed at the guitar: lowered in frequency so it catches the metallic zing of the acoustic's strings and some nasal vocal edge, not classic vocal sibilance. Use your de-esser's audition mode to hear exactly what it grabs.
- Scheps Omni Channel: one plugin doing light saturation for edge, a 90 Hz filter, a small dip at 413 Hz, and a gentle 8 kHz shelf. The shelf brightens vocal and guitar together, which is the whole game: find the amount of top end that flatters both.
- 1176 Blue Stripe, all buttons in, mix knob low: aggressive compression blended in at a whisper. Soloed you hear it breathe on the guitar; in the mix it just adds energy.
Parallels do what inserts cannot
Because heavy insert compression would wreck the guitar, the vocal gets bigger through parallel channels instead:
- A 610 tube compressor parallel bringing up body in the voice
- A SansAmp parallel (at the artist's request) for haze and fuzz
- The Scheps "vocal air" parallel: cut lows and boost highs into an LA-2A limiting high frequencies, then restore the lows, for a thin, airy brightness that helps lines cut through without brightening the acoustic
Parallels are also easy to ride: automate the parallel fader up on lines that need help and the core sound never changes.
Space that flatters both
The track feeds a tight tape-echo slap (under 100 ms, from an Echoplex) that thickens the voice and gives it a room, and a live chamber delayed by 90 ms so the reverb blooms after each word instead of on top of it. Both effects read as "vocal effects" but flatter the strumming too.
FAQ
Can AI stem separation split the vocal from the guitar instead?
It can, but you lose what made the recording special: the blend, the bleed and the commitment. For records with this aesthetic, mixing the combined track usually sounds better than separating and re-gluing it.
How do I EQ a track with two instruments on it?
Find the moves that help both (usually broad, gentle top and bottom shaping) and keep problem-solving cuts narrow. Audition everything in the full mix.
Should I ever record vocal and guitar to one track on purpose?
If you perform them as one thing, yes. Committing forces better decisions at the source and captures a performance no overdub reproduces.
Watch Matt work through this entire track, parallel by parallel, in the new Inside the Mix series on Dr. Dog's "Fat Dog," filmed at Southern Grooves in Memphis.
