Yangqin / Chinese hammered dulcimer multi-mic phase: courses + hammers vs soundboard, room/OH, and Chinese ensemble fold
Close spots on yangqin courses and bamboo hammers feel huge in the booth and in cans. On phone speakers or YouTube mono fold, hammer attack, Mid, and soundboard bloom often thin or cancel against stereo OH and room paths that carry the same phrase a few milliseconds later. Chinese ensemble beds (erhu, pipa, guzheng, dizi, voice) stack the same problem. That is usually polarity and arrival time between course/hammer spots and shared beds, plus a stereo bed that was never folded before send. Fix polarity and small delays first. Frequency-aware phase repair belongs only after residual L/R fights remain on stereo beds before mono fold. Distinct from Hungarian cimbalom, mountain dulcimer/zither, and Persian santur: this page is about Chinese yangqin alone. Cross-links: cimbalom, dulcimer/zither, guzheng, pipa, guqin, field, livestream.
Download PHIXER free Learn more hub
PHIXER is free stereo phase-repair (VST3/CLAP/AU · Win/Mac/Linux) from Riverflux Audio. Frequency-aware phase rotation for leftover L/R fights on stereo material. It is not a compressor, not a widener, and not a substitute for mic placement or mono audition of the stream or mix bus. Polarity and timing first; PHIXER after residual spectral fights.
What Chinese yangqin hammered-dulcimer multi-mic phase failure sounds like
- Course / hammer bite huge in cans, thin on phones. Close course or hammer spots share the same phrase a few ms apart from stereo OH or room.
- Soundboard bloom hollow on YouTube. Spot + stereo bed + room stack cancels Mid when the platform folds.
- Bamboo-hammer attack present in booth, gone on phone. Hard L/R across left/right course and bridge paths collapse on mono fold.
- Yangqin hero vs erhu / pipa / guzheng / dizi / voice bed fight. Chinese ensemble beds and yangqin spots share mid energy without polarity hygiene.
- Fine in headphones, broken for livestream viewers. Same destinations as YouTube / social mono, OBS livestream, cimbalom, dulcimer / zither.
Fast mono audition (about 60 seconds)
- Solo the stream / mix bus (not only booth headphones). Flip a utility to mono.
- A/B yangqin course/hammer spots muted vs open under stereo OH and room. Note what disappears: hammer bite, Mid, soundboard bloom, bridge body.
- Solo pairs: left vs right courses, hammer close vs soundboard secondary, yangqin vs erhu/pipa/guzheng bed, spot vs OH.
- Check correlation on the delivery bus, then with the full Chinese ensemble bed muted vs open.
Order of operations (before spectral phase rotation)
1. Polarity between course/hammer spots and shared OH / room / soundboard
Flip course or hammer spots vs stereo OH, or yangqin vs companion bed, when body or Mid return immediately on a clear phrase. Guide: polarity vs phase. Cousins: cimbalom, dulcimer / zither, guzheng.
2. Small delay align secondary paths to the hero yangqin image
OH, room, and soundboard secondary often arrive later than close course/hammer spots. A few samples of delay on the secondary often restores body better than EQ for "more yangqin."
3. One hero yangqin image to the stream or mix bus
Do not send every course pair, hammer close, soundboard, room, and OH all hot into the livestream or print. Pick a hero Mid-safe bed (often a focused stereo OH plus selective course spot), keep extras muted or low. Parallel stacks are classic thin-yangqin on phone fold. Related: overhead & room.
4. Prefer hero courses as the tone; quieter color spots
Secondary soundboard and bridge closes should color proximity and board identity, not replace the hero image for the stream mix. If every close sits at full level under a loud OH, mono fold almost always hollows the yangqin bed.
5. Skip wideners and Haas on yangqin beds before fold
Short-delay width and stereo imagers make booth headphones shimmer and phones hollow. Keep critical low-mid and soundboard bloom centered for stream and print.
6. Only then audition residual L/R spectral fights
If a stereo bed must stay stereo and fold still hollows a band after the steps above, gentle frequency-aware phase rotation can help. Prove it with mono A/B, rotator bypassed vs engaged.
Path types and typical failure modes
| Path | Common mono failure | First fix |
|---|---|---|
| Course/hammer spots + stereo OH | Body / Mid cancel on stream | Polarity; delay secondary; quieter secondary |
| Yangqin spot + soundboard + room | Hollow hammer attack, thin board bloom | Polarity + timing; hero Mid-safe path |
| Yangqin + erhu / pipa / guzheng / dizi bed | Hero vs companion fight on phone | Polarity + delay; hero image |
| Stereo pair hard L/R across left/right courses | Huge in cans, thin on phone | Center critical lows; simplify width |
| Board / matrix → OBS / YouTube | Fine in booth, hollow for viewers | OBS livestream / YouTube mono |
Studio / livestream checklist (to send)
1. Label courses vs hammers vs soundboard vs OH vs room vs companion bed
Name left/right course spots, hammer closes, soundboard secondary, erhu/pipa/guzheng/dizi/voice, stereo OH, and room so operators know what hits the livestream or print. Ambiguous stacks get left open and comb on fold.
2. Mono-check delivery bus alone, then with full Chinese ensemble bed
Solo hammer body and Mid in mono. Then open soundboard bloom and course bite in mono. Note fights between secondary color and hero bed.
3. Watch phone, YouTube, and social destinations
Most remote viewers hear mono or near-mono. Guides: YouTube / social mono, OBS livestream, TikTok / Reels / Shorts.
4. Re-check after stand / seating / angle moves
Instrument angle and mic height change distance and bleed. Flip polarity again after you move stands, change OH height, or re-seat the player.
When PHIXER helps
- The material is already stereo and must stay stereo (stereo OH, stereo room or board print), not a mono close you fake-widened.
- Polarity, delay align, and hero-path levels are cleaned up, but fold still hollows a band of the spectrum.
- You can A/B mono with the plugin bypassed and hear clearer body or Mid without wrecking the image.
- You are not using it to "fix" unaligned course+OH under room, or to replace a time-align edit on the matrix.
Decision table
| You notice… | First try | Why |
|---|---|---|
| Body thins when OH opens under course/hammer spots | Polarity; delay secondary; quieter secondary | Arrival-time fight |
| Yangqin + companion harsh and hollow | Polarity; timing; hero Mid-safe path | Dual-path spot + companion |
| Fine in booth, thin on phone | Mono delivery bus; pull extras | Booth ≠ viewer fold |
| Stereo bed still cancels by band | PHIXER on that stereo bed | Residual spectral L/R fight |
| Mono yangqin alone is dull | Tone / EQ / gain / mic choice | Not an L/R repair job |
When NOT to use PHIXER here
- You have not mono-auditioned the stream / mix bus yet.
- The fix is obviously polarity, delay, one hero image, or quieter secondary under course/OH.
- You want PHIXER to make the yangqin louder, wider, or "more Chinese ensemble." Wrong job.
- The path is mono and the print problem is balance or EQ.
- See also when not to use PHIXER.
Quick path: Solo delivery bus in mono → polarity course/hammer spots vs OH/room/soundboard → delay-align secondary → one hero image → quieter extras under hero → skip Haas/imagers → mono-check with full Chinese ensemble bed → PHIXER only for leftover stereo cancellation on a stereo bed → fold again for stream / studio print.
Related guides
- Cimbalom / Hungarian folk multi-mic phase (distinct: Hungarian cimbalom vs Chinese yangqin)
- Santur / Persian hammered dulcimer multi-mic phase (distinct: Persian santur vs Chinese yangqin)
- Hammered dulcimer / mountain dulcimer / zither multi-mic phase
- Qanun / Middle Eastern zither multi-mic phase
- Guzheng / Chinese zither multi-mic phase
- Pipa / Chinese lute multi-mic phase
- Guqin / Chinese seven-string classical multi-mic phase
- Field recording / dual recorder outdoor phase
- Polarity vs phase rotation
- When not to use PHIXER
- YouTube / social mono fold checklist
- Twitch / OBS livestream mono fold
Watch
- PHIXER tutorial + how it sounds (walkthrough with audio examples)