A Song 7×7 viola d’amore is a good hybrid candidate precisely because the sympathetic bank is already a physical resonator. Reset the nut, bridge, and soundpost first, then add electronics that treat the bowed strings and the resonance strings as two sources. Do not drill the top until a clamp-on system has proved the idea.
Song workshop instruments of this type are typically a 14-inch body with about a 13-inch string length, seven playing strings over the bridge and seven thin steel or brass sympathetics passing through the bridge and under the fingerboard. Common open tuning is A–d–a–d′–f♯′–a′–d″, with the sympathetics tuned in parallel or, if you prefer, as a scale. Listings often ship with the wrong strings, so restring before you judge either the acoustic setup or a pickup.
Reset the acoustic path first
Electronics will amplify whatever the setup is doing.
- Bridge. Feet must sit flush on the arch, body perpendicular to the top, bass side slightly higher. Playing strings sit in shallow notches on the crown; sympathetics pass through clean holes below, free of the fingerboard underside and not buzzing on the hole edges. A buzzing sympathetic will dominate a piezo.
- Soundpost. Just behind the treble foot, visible through the flame hole, upright and fitted. Move it in small steps. On these instruments it balances the silvery upper partials against the lower playing strings; a post that is too tight kills the sympathetic bloom you probably want as a synth source.
- Nut. Two problems, not one. The playing-string nut needs even spacing and low, clean action. The sympathetic path through the pegbox and under the fingerboard needs clearance so those wires do not tick against wood. Song nuts are often rough; a recut nut and a light dressing of the bridge holes usually do more than any preamp.
Use real sympathetic wire (thin steel on the top three, brass on the lower ones is the usual split) and playing strings meant for the instrument. Gut or synthetic playing strings sound right acoustically and will not drive a magnetic pickup. Steel or metal-wound playing strings will.
Three electronic layers, least invasive first
1. Whole-instrument bridge piezo. A violin/viola bridge-eye piezo (KNA VV-style, or a small disc under one or both bridge feet) hears both banks mixed through the body. Passive piezos need a high-impedance preamp — a Radial PZ-DI, Fishman, or similar, with the orchestral/piezo load setting. Clamp the jack to the lower bout. This is the practical stage path: acoustic blend in one channel, effected signal in another.
2. A second transducer only on the sympathetic bank. This is the modification that makes the instrument interesting rather than “a viola with a pickup.” A small contact mic or thin piezo near the lower block, tailgut, or the bridge holes where the sympathetics terminate gives you a channel that is mostly ring and almost no bow noise. That channel is the one to send to granular, convolution, or a long resonant filter. The bowed channel stays the melody.
3. Pitch and envelope into a synth, not a hex pickup. Seven courses plus a resonance bank will not track cleanly on guitar-style MIDI pickups. Use the bowed piezo into a monophonic or lightly polyphonic tracker (Jam Origin, Max, Ableton, hardware like a Sonuus for single lines). Drive oscillators from pitch, and use bow energy as an envelope follower so dynamics still come from the stick. Mute or detune the sympathetics when you need clean tracking; open them when you want the tracker to follow a cloud instead of a note.
Magnetic pickups only make sense if the playing strings are steel. A small magnetic element under the fingerboard can hear the sympathetics alone, since those are already wire. That is an experiment, not a first install.
Signal path that actually uses the instrument
A useful hybrid chain looks like this:
- Channel A, bridge piezo: dry or lightly compressed, your audible viola d’amore.
- Channel B, sympathetic contact: high-pass a little, then convolution, granular freeze, or a resonant low-pass that follows Channel A’s pitch.
- Channel C, synth: pitch and amplitude from Channel A, with Channel B as the reverb or the noise source the synth is filtered by.
Cross-synthesis is the natural trick here. The sympathetic strings already do physical convolution; an FFT convolve of the bowed signal with a synth pad, or of the sympathetic channel with a sampled impulse, extends that instead of replacing it. Research instruments such as the EV electronic viola use exactly this — pickup for pitch and amplitude, then convolution of the acoustic signal with the synth — and recent work has even driven a viola d’amore’s wire strings with magnets and current (Lorentz force), so the sympathetics can be actuated as well as heard. That last step is a lab build, not a weekend mod.
Keep the acoustic top intact. Clamp-on jack, bridge-eye or foot piezo, one extra contact on the sympathetic termination. If you later want a permanent endpin jack, a luthier can fit it at the bottom block without opening the resonance you just set. The soundpost and bridge you reset are doing the mixing; the electronics should split that mix, not fight it.
Granular synthesis is the right processor for the sympathetic channel on that viola d’amore. The bank is already a cloud of decaying partials; granulation just lets you decide how long a moment of that cloud lasts, how wide it spreads, and whether it stays tied to the bow.
A grain is a short slice of audio, usually 5–200 ms, faded in and out so the splice does not click. Grains are read from a buffer and overlapped. When they overlap enough, the ear stops hearing fragments and hears a texture. Curtis Roads’s working range is about 1–100 ms; under about 20 ms the slice is too short to carry a clear pitch, and over about 80–100 ms you start hearing the source as a micro-loop rather than a wash.
The parameters that matter
Six controls do almost all the musical work.
- Size. Short grains (5–15 ms) click, buzz, and lose the note. Medium grains (30–60 ms) keep the silvery partials of a sympathetic string and still blur the bow. Long grains (100–400 ms) preserve attack, vibrato, and the ring itself.
- Density. Grains per second. Sparse (4–8/s) is pointillist; dense (40–100/s) is a pad. Overlap is roughly size times density. A 50 ms grain at 40/s is a heavy overlap and sounds continuous.
- Position. Where in the buffer the read head sits. Scanning forward is time-stretch. Scanning at zero is a freeze.
- Spray / jitter. Random offset around position, pitch, or pan. Freeze without spray is a loop. Freeze with a few milliseconds of spray breathes.
- Pitch. Per-grain transposition. A tight random spread of ±5–15 cents is chorus. Octaves and fifths turn one bowed note into a chord cloud.
- Envelope. Raised cosine or Gaussian fades are smooth. A square window clicks. Shorter fades on longer grains sound more plucked.
Pitch and time separate because each grain is replayed at its own rate while the position pointer advances at another. Slow the pointer, keep grain pitch at 0, and the note holds without dropping. That is the whole trick.
Techniques, with settings that suit this instrument
Freeze. Stop the position pointer on a ringing sympathetic transient — the moment just after the bow leaves a double stop is ideal. Size 60–80 ms, density high, spray 5–20 ms, pitch scatter off or ±7 cents. You get a held chord of whatever the bank was resonating, and the bow can move on. This is the one to map to a footswitch.
Cloud. Same freeze region, but open spray on position, pitch, and pan. Pitch scatter of a few semitones, not cents, turns the d’amore’s silvery tail into a choir. Keep one dry acoustic channel underneath or the attack disappears.
Time stretch of the ring. Play a phrase into the buffer. Advance position at 10–25% speed, grain pitch fixed, size around 80 ms. The decay that normally lasts two seconds lasts eight, and the tuning does not fall. Good for writing lines the acoustic instrument cannot sustain.
Pitch smear without a tracker. Quantize grain pitch to the open tuning, A–d–a–d′–f♯′–a′–d″, or to a just fifth and octave. A monophonic bowed note then sprays only onto notes the instrument already owns. It sounds like the sympathetics woke up late, which they essentially did.
Rhythmic granulation. Size 20–40 ms, density locked to a clock, spray on position only. The resonance bank becomes a soft sequenced echo of whatever you just bowed. Low density so each grain is a separate event.
Reverse grains. A fraction of grains play backward. On a bowed decay this pulls the swell out of the tail and puts it at the front of the texture. Use sparingly; 10–20% reverse is enough before it turns into a cartoon.
Adaptive size. Shorten the grain when the buffer is noisy (bow scratch, string change) and lengthen it when the buffer is a stable harmonic. The texture stays smooth on sustains and does not mash the attacks into mud. Most of the better software engines can do this from an envelope follower on the input.
Grain delay, not a synth. Very low density, size matched to the note, feedback inside the granulator. This is closer to a multi-tap delay whose taps are slices of the sympathetic tail. It sits in a mix more politely than a full cloud.
Cross-granulation. Buffer A is the bridge piezo (bow noise, fundamental, articulation). Buffer B is the sympathetic contact. Read position from A’s envelope but take the audio from B, or the reverse. The bow’s rhythm drives a texture made only of ring. That is the hybrid move: the two pickups from the earlier setup become the two buffers.
Routing on the Song
Keep three paths.
The bridge piezo stays mostly dry, or only lightly compressed, so the instrument is still identifiable. The sympathetic contact feeds the granulator, high-passed a little so bow thump and body knock do not become grains. A third send from the bowed channel can duck or open the cloud: loud bow closes the grain level, end of the stroke opens the freeze. The sympathetic strings already do physical convolution; the granulator should extend that tail, not replace the note.
Capture a few seconds of open-string resonance with the bow muted and store it as a permanent buffer. Live input then only chooses position and density. The instrument can be put down and the cloud still plays, which is useful if you also want a synth layer triggered from pitch tracking — the tracker follows the dry bow, the granulator does not have to.
Where to do it
In a DAW, Ableton’s Granulator II, Output Portal, or Native Instruments Straylight cover freeze, spray, and pitch quantize. Max or Pure Data is better if you want the sympathetic contact and the bridge piezo as separate live buffers with a footswitch freeze. Hardware that behaves well on a bowed feed: Mutable Instruments Beads (or Clouds), Instruo Arbhar, and Make Noise Morphagene. Beads is the practical one — live input, freeze, and a density/size balance that does not need a laptop on stage.
Two failure modes are specific to this source. Grains under 20 ms throw away the partials you reset the soundpost to get. And a wide unquantized pitch spray fights the just intervals of the sympathetic bank; the acoustic instrument is already in tune with itself, so random semitones sound like a second, broken instrument rather than an extension of the first.
A microphone is often the better first transducer on a viola d’amore, especially if the point is the sympathetic ring rather than loud-stage isolation. A piezo is not required. It is the feedback-resistant option, not the more faithful one.
A clip-on condenser hears air. It takes the top, the flame holes, the bow, and the silvery decay of the wires under the fingerboard. A bridge piezo hears pressure in the wood. It takes attack and body vibration, rejects the room, and mostly misses the air resonance that makes this instrument different from a viola. Most “contact mics” sold for instruments are themselves piezoelectric discs glued or puttied to the top, so they sit between the two: more body than a condenser, less bow detail, and still prone to thumps.
For the hybrid path we discussed, the useful split is a clip-on for the live acoustic image and, only if you need a second clean channel, a piezo or contact on the bridge or lower block for the granulator. One mic can do both jobs at home. On a loud stage it will howl before the piezo does.
Fitting a microphone
Nothing is cut. A small cardioid condenser on a clip or gooseneck is enough: Audio-Technica Pro 35, DPA 4099 with the violin mount, or a cheap violin clip-on if you are only testing the idea. Phantom power or a battery pack is required; a piezo jack will not power it.
Clip the mount to the tailpiece, not the bridge. The bridge is already carrying seven playing strings and seven wires through its holes, and a clip there dampens it and buzzes against the sympathetic holes. Point the capsule at the treble flame hole or just above the top playing strings, about 1–2 cm off the wood, aimed slightly downward. That spot hears the hole and the string crossing without sitting in the rosin cloud.
If the tailpiece clip slips, the lower bout edge or the chinrest bracket is the next mount. Do not tape a capsule over a flame hole. You choke the air resonance and the mic hears a honk. Keep the cable along the bout and out past the shoulder so it never crosses the bow.
The same mic can feed the dry channel and the granular buffer. For a dedicated sympathetic feed, putty a second small capsule or contact under the tailpiece, facing the lower block where the wires end, and keep it off the top. High-pass that channel before the granulator so bow knock does not become grains.
Feedback is the limit. Cardioid, close placement, and a notch around 200–400 Hz if the body starts to ring. Wedges in front of the instrument will fight you; side fills or in-ears will not.
Fitting a piezo if you still want one
Use a removable violin/viola bridge-wing pickup, KNA VV-3 or Fishman V-100 style, not an undersaddle guitar pickup. The jack housing clamps to the lower bout on the bass side with cork-lined barrels. No hole in the top. Song bodies are viola-sized, so use the longer clamp barrels if the waist is too wide for the violin set.
The element presses into a wing slot of the bridge, bass or treble. Measure the slot. Fishman’s element wants roughly 1.5–2.3 mm; KNA’s wooden disc is about 1.8 mm thick. If the slot is tight, slacken the strings, lift the bridge, and widen only that slot with a fine file or sandpaper wrapped on a feeler. Stay inside the slot. Do not thin the feet or the crown, and do not open the sympathetic holes. If the slot is already wide, which is common on viola bridges, shim with a scrap of card on the leg face so the element cannot rattle.
The sensing face must bear on one wall of the slot and the spring on the other. A loose element buzzes in the same register as a sympathetic wire, and you will not be able to tell them apart. Refit the bridge perpendicular, feet flush, then bring the strings back up and check that the wires still pass the holes without ticking.
Plug into a high-impedance input, 1 MΩ or more: a Radial PZ-DI, Fishman preamp, or the piezo setting on an acoustic DI. A normal mixer line input makes it thin and quacky. Roll a little off around 1.2–1.5 kHz if the attack is harsh.
Leave the soundpost where you set it. The piezo does not replace that setup; it reports it. If the wing slot on this bridge is too narrow to file safely, or the sympathetic holes sit too close to the slot, putty a disc contact to the top just beside a bridge foot instead. It comes off with a thumbnail and still gives the granulator a body channel while the clip-on keeps the air.
Start with the clip-on mic and a granulator. Add the piezo and the DI only if you want a second, feedback-resistant channel for the sympathetic feed.
Clip-on microphone. This is the one that hears the air and the sympathetic ring. No bridge work.
- Audio-Technica PRO 35, cardioid condenser with a clip and phantom-power module: audio-technica.com
- DPA 4099 CORE+, the better violin/viola mount and the one that behaves on a louder stage: dpamicrophones.com
Clip either to the tailpiece, not the bridge, and aim it at the treble flame hole.
Bridge piezo, optional. Removable, clamps to the lower bout, element sits in a bridge wing slot. Buy this only for the granulator channel or for stage volume.
- KNA VV-3, the one to get: knapickups.com and, in the UK, Gear4music
- Fishman V-100, same idea, jack on the tailpiece: johnsonstring.com
- How the wing slot is widened if it is too tight: Electric Violin Shop
Piezo DI. Required if you buy the KNA or the Fishman. A normal mixer input makes a piezo thin and quacky. Set the load to 10 MΩ.
- Radial PZ-DI: radialeng.com
The microphone does not need this. It needs phantom power from the interface or mixer.
Granular processor. Software is the practical route unless you already have a Eurorack case.
- Output Portal, live input, freeze, and scale-locked pitch: output.com
- Ableton Granulator II, if you already have Suite. It is a Max for Live device, not a separate purchase.
- Mutable Instruments Beads, only if you have a modular case and want it on stage with no laptop: Reverb listing
Minimum that matches the plan: the PRO 35, an audio interface with phantom power, and Portal or Granulator II. The KNA plus the PZ-DI is the second purchase, once the mic channel is working.
For soundtrack and DAW work, a reset Song and a clip-on mic is the right pair. The piezo is the stage tool. You are not on a loud stage.
The reset is what makes the recording usable. A Song 7×7 with a fitted bridge, a soundpost that is not choking the top, and sympathetic wires that are not ticking will give you the silvery tail that survives compression, reverb, and a picture edit. A piezo reports the bridge and throws most of that air away. Record the mic. Keep the piezo, if you buy one later, as a second track for experiments.
Split the signal in the interface, not in the instrument.
- Track 1 is the mic, dry or lightly compressed. That is the soundtrack stem.
- A send from that track feeds Portal, Granulator II, or whatever pedal-plugin you already trust. Freeze on the decay after a double stop, not on the bow attack.
- A copy with the sympathetics damped, or a high-pass and a gate, is the track you turn into MIDI.
Audio-to-MIDI is the weak point, and the mic is not the cause. Seven playing strings plus seven ringing wires look like a chord that never ends. Trackers will spray ghost notes. Ableton’s Convert Melody to MIDI, Spotify Basic Pitch, and Jam Origin MIDI Guitar all do this. They cope with a single bowed line. They do not cope with an open sympathetic bank.
Damp the wires with a strip of felt under the fingerboard when you want MIDI. Play one string. Record that pass, convert it, then take the felt off and record the same line again for the soundtrack stem. Offline conversion on a clean take beats real-time tracking. Real-time is worth it only for a monophonic lead you are using to drive a synth while you play.
Pedals sit after the interface, not after the mic. The PRO 35 needs phantom power; a guitar pedal will not supply it, and the level will be wrong. Mic into the interface, then either a reamp box out to the pedalboard and back, or the pedals you already like as plugins on the mic track. Distortion, filters, and grain delay want the damped or high-passed copy. Reverb and delay can take the full mic, because the sympathetic ring is already the effect.
So the fit is good if you treat it as three jobs: the reset instrument and the mic for the score, a damped pass for MIDI, and pedals on a send rather than on the only cable. You do not need the KNA or the PZ-DI for this workflow. Buy those if you later want a feedback-proof live channel.
A reset Song is a usable workshop viola d’amore, not a specialist one. The setup closes the gap that makes these instruments arrive dead. It does not close the gap in the wood, the graduation, or the bridge voicing. For soundtrack work that is often enough. For a recital instrument it is not.
Song workshop 7×7 instruments list around $700–800. A specialist 7×7, Jonathan Hill’s for example, starts at £16,500. The Song is a 14-inch body, about a 13-inch string length, spruce top, flamed maple back, angel head, and the usual A–d–a–d′–f♯′–a′–d″ tuning. Listings admit the strings are wrong and have to be replaced. That is the whole difference a reset can fix: nut, bridge fit, soundpost, and real playing strings plus thin steel and brass sympathetics. After that it will ring, speak, and record. It will not have the evenness, the clean bass, or the controlled silvery top of an instrument whose top was graduated for fourteen strings.
Against a violin or a viola, the Song is quieter and slower to speak, and the note you bowed is never alone. A violin gives you attack, projection, and a clean line that cuts a mix and tracks to MIDI. A viola gives you the alto register without the extra wires. The Song gives you a halo that lasts after the bow leaves. In a soundtrack that halo is the point. In a rhythmic cue or a MIDI pass it is the problem, which is why the damped take still matters.
Against a Hardanger fiddle it is the same idea and a different job. The Hardanger has four or five playing strings and four or five sympathetics, a flatter bridge, and a folk repertoire with scordatura. It is a little quieter than a violin and more intense in the ring. The viola d’amore is larger, has six or seven bowed strings, and was built for chordal Baroque writing rather than dance tunes. On a recording the Hardanger reads as Nordic and bright. The Song, once set up, reads as darker and more choral. Neither replaces the other.
A nyckelharpa is the other sympathetic cousin, and it is worse for your workflow. Keys instead of a fingerboard, drones by design, and a chopping bow. It makes a strong texture and a poor melody-to-MIDI source. A sitar has the same physical trick, a bank of wires that ring, but the attack is plucked and the tuning is not the one your pedals and scale-locked granulator already expect.
For the work you described, the reset Song sits between a viola stem and a pad. It will not outplay a good viola on a solo line, and it will not track as cleanly as a damped violin. It will give you a bowed source that already sounds like the granular freeze you would otherwise build, and a $800 workshop body is the right thing to clamp a mic to and run through pedals. The specialist instrument is for the player who needs the real repertoire to speak evenly. You need the ring. The reset is what makes the Song produce it.
It will not give you a new range. Violin, viola, cello, and bass already cover that. What it adds is a bowed instrument whose open tuning and sympathetic bank do something none of those four can.
Your viola occupies the same register and will always be the better solo line. It speaks faster, tracks to MIDI more cleanly, and sits in a mix the way an editor expects. The d’amore, once the nut, bridge, and soundpost are actually set, is tuned A–d–a–d′–f♯′–a′–d″. That is a D-major chord across the open strings, not fifths. A simple bow stroke is already a chord with a metal-string halo that keeps ringing after the bow has left. Your viola cannot do that without you faking it with a send. Your violin is too bright and too clean. Cello and bass can drone, but they do not produce that silvery, late upper partial.
For the work you described, that is the interesting part. A soundtrack bed, a granular freeze, a pedal reverb that already has a source longer than the note: the Song is a texture instrument sitting between the viola and a pad. Scordatura is normal on it, so retuning the sympathetics to the key of a cue, or to a scale instead of parallel with the playing strings, changes the halo without a plugin. Double stops that are awkward on the viola fall under the hand. The line you would bow on the viola can be recorded again on the d’amore and layered under itself, and the two will not phase like two violas.
It will not enlarge the repertoire in the classical sense. Almost nothing written for it is unavailable on viola, and the Song will not be the instrument you take to a quartet. It also will not improve the MIDI path. The wires that make it useful are the wires that spray ghost notes, so the damped pass you already planned still stands.
Worth it if you want a bowed source that is already a chord and a tail. Not worth it if you want another melodic voice. You have those.
