Blog
Horia Stan6 min read

Tempo-Synced M/S Vocal Textures: My 2026 Chain for Stereo Motion That Stays Mix-Ready

A 2026 vocal chain using tempo-synced M/S delays, formant-locking and dynamic multiband side processing to create durable stereo textures.

Horia Stan is a music producer and sound engineer at The One Records, Bucharest.

What this is and why it matters

I make vocals sound like an instrument and not a mess. That means a dry center that cuts and a side-field that breathes rhythm, movement and colour. Most producers either widen everything and lose focus or keep the vocal mono and boring. My approach uses M/S routing plus tempo-synced modulation and dynamic side processing so the stereo motion is musical, mono-compatible and stays intact after streaming codecs and mastering.

I run this inside Logic Pro. I monitor on an Audient iD14 MkII and I use FabFilter Pro-Q3 and Pro-MB as the dynamic backbone. Waves Doubler or H-Delay show up as flavour tools. Keyscape pads sit behind the textures when the arrangement needs density.

Make the sides do the movement - not the mid. Control that movement with the vocal envelope so the stereo field grooves with the singer, not against them.

High-level chain

  1. Lead vocal track - dry comp and air. Keep this mono. Use Logic’s Compressor or FabFilter Pro-C2 for transparent control.
  2. Bus 1 - Mid bus - mono sum of the vocal with surgical EQ and de-essing. This is the centre.
  3. Bus 2 - Side rhythm bus - stereo processing only. Tempo-synced delays, micro-delays, formant-shift on sends, and multiband dynamic side compression keyed to the vocal envelope.
  4. Bus 3 - Texture bus - long convolution or reverb processed into a rhythmic pad via gating and LFOs. This is optional when the mix needs width without masking.

Exact plugin settings I use (start here)

Lead (raw)

  • Logic input chain: Gain staging so peaks hit -6 dBFS. I record to 24-bit. Audient iD14 MkII at buffer 128 samples for tracking.
  • EQ: FabFilter Pro-Q3 - highpass 80 Hz, Q 0.7. Cut 2.5 dB at 250 Hz if boxy. Boost 3.5 kHz +2.3 dB with a bell Q 1.2 for presence.
  • Compression: FabFilter Pro-C2 set to 3:1, 8 ms attack, 60 ms release, 3-4 dB gain reduction on peaks.

Bus 1 - Mid (centre) - surgical

  • Route the lead to a Bus labelled 'Vox - Mid'.
  • Insert Pro-Q3 in M/S mode. In the Mid channel: HPF 80 Hz, narrow cut at 3.5 kHz if harsh. Keep this centred and dry.
  • Output target: aim for -6 to -8 dBFS peak on the bus.

Bus 2 - Side rhythm - the creative bus

  • Send the lead to 'Vox - Sides' at -6 dB send level as starting point.
  • On the bus, insert an M/S encoder if needed. Or use Pro-Q3 M/S focusing processing on the Sides only.
  • Step 1 - Micro-width: Insert a micro-delay plugin. Logic’s Stereo Delay or Waves H-Delay works. Set Left delay = 11 ms, Right delay = 23 ms, feedback 0, lowpass 6 kHz, hi-pass 200 Hz. This gives spatial smear without obvious echoes.
  • Step 2 - Tempo layer: Add a tempo-synced delay on a parallel send inside the Sides chain. At 120 bpm, use a dotted 1/16 = 187 ms for slow groove hits. Use feedback 15-25% for 2 repeats, lowpass 4.5 kHz. Automate send level per section to avoid wash.
  • Step 3 - Formant motion: Insert Waves Vocal Bender or Soundtoys Little AlterBoy on the Sides insert. Set formant shift to +/- 0.8 to 1.2 semitones and mix to taste 20-40%. Automate the formant on pre-chorus fills to create movement. Because this runs only in Sides, the mid stays intelligible.
  • Step 4 - Dynamic side shaping: FabFilter Pro-MB on Sides only. Create 2 bands: 200-800 Hz and 2.5-6 kHz. Set the 200-800 Hz band to look like a compressor with external sidechain coming from the original vocal's RMS. When the vocal hits, the Sides duck - ratio 2.5:1, attack 6 ms, release 90 ms. For 2.5-6 kHz, use soft compression 1.8:1 so sibilance in the sides breathes but does not clash.

Bus 3 - Texture (optional)

  • Long reverb or convolution. I use a plate IR for darker textures.
  • Pitch-shift the send -3 to -7 cents for thickness, or full semitone for subtle detune. Lowpass at 6 kHz. Gate the reverb with an LFO synced to 1/8 or 1/16 so it reads rhythmically.

Automation and performance tricks

  • Envelope follower to tempo map: Use Logic’s sidechain or an envelope follower plugin to modulate the left-right micro-delay split. When the vocal verse is sparse, widen the delay split by +6 ms left/right. When the chorus hits, reduce to 3/7 ms so the vocal stays gluey.

  • Section gains: I print two stems for mastering - 'Vox Dry' and 'Vox Sides'. I deliver these with the same LUFS target as the mix. I aim for the combined vocal to sit ~ -14 LUFS integrated on the mix bus before mastering. Adjust the side bus to taste - more side energy raises perceived loudness but can fight the lead.

  • Mono check: Collapse to mono after applying the Sides chain during mixing. If comb filtering kills the vocal, change micro-delay timings by +/- 2 ms or reduce the formant mix. The goal is movement that vanishes gracefully to mono, not phase chaos.

1
Build the three buses
Create Mid, Sides and Texture. Route the lead via sends. Keep the lead track as the source of truth.
2
Side processing order
Micro-delay first, tempo-synced delay second, formant third, then Pro-MB dynamic shaping.
3
Freeze and test
Bounce the Sides to a stereo file and test both channels summed to mono, and run it through export codecs if you must.

Why this works - short technical logic

  • Center intelligibility stays because the main vocal remains un-modulated in Bus 1.
  • The Sides create rhythmic motion because tempo-synced delays lock to the pocket. Micro-delays provide width without obvious echoes.
  • Formant shifts on the sides give a sense of other-worldliness without changing the singer’s timbre in the mix centre.
  • Multiband side compression keyed to the vocal envelope ensures the side-field breathes with the performance and never overtakes the lead.

Common mistakes I see

  • Applying the same chorus or doubler to the lead and the sides. That blurs focus.
  • Using long reverb on the lead and expecting clarity. Put long tails in the sides or texture bus, not the mid.
  • Overdoing formant shift on the mid. Small amounts on the sides are far more musical.

When to skip this

If the song needs pure vintage mono vocal or a raw, close intimate production, skip the sides chain. This method is not for lo-fi singer-songwriter arrangements. It is for modern dark pop and cinematic pop where the vocal must be both center-right and atmosphere-building at the same time.

Concrete takeaway - a checklist you can copy

  • Track dry vocal to -6 dBFS and compress 3-4 dB peaks with Pro-C2.
  • Route to three buses: Mid, Sides, Texture.
  • Sides chain: Micro-delay L=11 ms R=23 ms; Tempo delay dotted 1/16 at session BPM; Vocal Bender formant +/-0.8 semis at 20-40% wet; Pro-MB duck band 200-800 Hz ratio 2.5:1, attack 6 ms, release 90 ms.
  • Mono-check and adjust micro-delay by +/-2 ms if phase issues appear.
  • Export two stems: 'Vox Dry' and 'Vox Sides' for mastering or alternate mixes.

I use this on records I produce at The One Records. If you copy the exact numbers above you will save hours of chasing phantom width. Apply it to a verse first, then scale the send levels into the pre-chorus and chorus. That is the quickest way to add professional stereo motion without killing the vocal.

vocal-productionMS-processingLogic-ProFabFilter2026-workflow