The studio shelf holds objects that are older than the notation we write for them: a copper bowl that rings for a minute after one strike, a gong that bends downward in pitch as it decays, a frame drum, a conch, a rattle made from a dried gourd. They sound like nothing digital can imitate, which is why they are in the room. The trouble starts when the same objects are sold as medicine.

This is the sorting I do before a session: what the instrument physically does, what the trials actually measured, and what belongs to the person or tradition it came from. None of that requires a miracle claim, and all of it makes the instruments easier to use well.

Singing bowls: the physics is settled

A Himalayan bowl is bronze — copper and tin, in the same family as a bell — and it sounds the way it does because the wall vibrates in natural modes. Terwagne and Bush's 2011 paper in Nonlinearity is the standard account: rubbing excites the modes through stick-slip friction, striking excites the same modes impulsively, and with water in the bowl the vibration drives surface waves and can throw droplets into the air. It is a beautiful paper about a resonant object. It makes no health claims at all.

The popular metallurgy is less reliable. Vendors describe bowls cast from “seven sacred metals”, a list that traditionally includes lead and mercury. No peer-reviewed assay I can find supports the recipe, and taken literally it describes an object you would not want against skin. Alloy content in unregulated small-scale casting is unpredictable. If a bowl matters to your practice, buy from a maker who states the alloy, and treat the seven-metals story as folklore rather than a specification.

What the bowl studies show

Two systematic reviews frame the field. Stanhope and Weinstein's 2020 review found only four peer-reviewed studies of singing bowls, two of them with weak designs, and concluded that singing bowl therapies could not be recommended at that stage. A 2025 review by Lin, Yang and Wang found fourteen studies and reported potential for reduced anxiety and depression and improved heart-rate variability — but it could not pool the results, because the studies were too different to combine, and it calls for rigorously designed controlled trials.

The best single trial so far is Fernández-Teruel and colleagues, 2023: fifty adults selected for high state anxiety, one forty-five-minute session, three arms — Tibetan singing bowls, progressive muscle relaxation, and a waiting list. Bowls reduced alpha power and increased heart-rate variability against both comparators, and both active arms reduced self-reported anxiety, with the effect more evident in the bowl group. The limits are the usual ones and they matter: one session, no active-attention control matched for the facilitator's presence, a waiting list that inflates every effect, and no possibility of blinding.

The observational study people cite most often is Goldsby and colleagues, 2016: 62 participants, one session, no control group, tension and anxiety down significantly. With no control arm it cannot show that the bowls caused anything. The detail I find most telling is that people new to sound meditation improved far more than experienced practitioners — which is exactly what novelty and expectation look like in a before-and-after design.

None of that makes bowls useless. It makes them an attention technology with a plausible relaxation effect and an evidence base that is encouraging rather than conclusive. That is a good thing to be.

Gongs: texture, not treatment

I could not find a single randomised trial of gong meditation. What exists is survey material — audience satisfaction from gong baths — and studies of multi-instrument sound baths in which gongs were a small fraction of the playing time. The acoustics are another matter: gongs are inharmonic, their partials do not line up in neat whole-number steps, and the pitch glides after the strike. That instability is the reason a gong can change a room in four seconds and the reason it is so hard to record well.

So I use gongs for what they demonstrably do: dynamics, transition, a physical event that resets attention. I do not describe a gong bath as a treatment, and I would be cautious with anyone who does.

The didgeridoo: the one real trial, and what it proved

Puhan and colleagues published a randomised controlled trial in the BMJ in 2006: 25 people with moderate obstructive sleep apnoea, 14 learning the didgeridoo and 11 on a waiting list, four months. The didgeridoo group's apnoea-hypopnoea index fell by 6.2 events per hour and daytime sleepiness by 3 points on the Epworth scale, with partners reporting less sleep disturbance. Sleep quality itself did not improve.

The mechanism is the part the wellness industry skips: practising the instrument trains the upper-airway dilator muscles. The didgeridoo worked as an exercise device, not as sound healing. The trial was small, the control was a waiting list because a sham didgeridoo is hard to design, and in twenty years it has not been independently replicated; a 2024 single-arm feasibility study is the closest follow-up.

You will also read that the didgeridoo has been used for healing for 40,000 years. That figure appears in the peer-reviewed literature, which is not the same as being archaeologically defensible. I do not repeat it.

Drums, rattles and conch horns

The claim that shamanic drumming at four to seven hertz drives the brain into theta has been repeated since the 1960s, when Neher proposed it as a hypothesis. A 2025 review of the neurobiology of rhythmic sound is careful about what has actually been shown: the strongest field evidence comes from a single participant in a ritual; the best imaging study used experienced practitioners and could not separate rhythm from ritual context; and the effects that do appear show up mainly in highly susceptible people. The review's own summary is that rhythmic sound facilitates absorption and relaxation, with entrainment still a proposed mechanism awaiting robust controls.

What rhythm reliably does is organise attention and coordinate a room. That is enough to build a session on. The conch and the pututu belong to the same story from the other direction: at Chavín de Huántar, the gallery network behaves like coupled rooms, so a horn played in one passage is heard somewhere else. The architecture does the effect; the instrument is the trigger.

Tuned stone and bronze

What these traditions share, beyond ritual, is precision. Tuned stone chimes and bronze bell sets were built as instruments, not as noise-makers: the bells of Marquis Yi of Zeng, buried around 433 BCE, each give two distinct pitches depending on where they are struck. Somebody designed those profiles to produce specific intervals, and the objects prove it. That is the part of the lineage I find most usable in a studio — the discipline of a tuned object — and it needs no metaphysical claim to be impressive.

Using them now, honestly

  • Play to their strengths: room activation, transitions, tactile low end, decay. A bowl after a digital track is a change of physical state that no plugin reproduces.
  • Record them properly and credit the maker. A close stereo pair, a quiet room, and the instrument's name and origin in the session notes.
  • State the alloy and the source when you sell a session or a sample pack. Where the material is protected — rosewood, for instance — finished instruments have had permit exemptions since 2019, but raw stock and non-instrument items are still controlled.
  • Keep unknown-metal bowls off skin and away from children. If the alloy is unstated, treat it as decorative rather than therapeutic.
  • Do not sell ceremony. Using a frame drum and a low drone is craft. Naming an improvisation after a specific people's protocol, or charging for an initiation you have not received, is something else.
  • Name the teachers, makers and communities you learned from. Attribution is the cheapest form of respect available and it costs nothing in a delivery sheet.
  • Keep the claims modest on the page: sound for relaxation, creative practice and ritual — not medical treatment.
  • If a collaborator needs a clinical claim, say plainly that these instruments do not carry one. Offer the experience instead, and let them decide.

What I tell a collaborator

These instruments are in the room because they produce a physical event that a room can feel — a strike, a decay, a beat between partials, a low frequency you notice in the chest before the ears. People relax in a well-held room, and bowls and gongs are excellent at holding one.

Beyond that, the honest position is narrow and durable: some trials show relaxation and mood benefits from sound-bath formats; one old trial shows a didgeridoo can train airway muscles; nothing shows that a gong or a bowl treats disease. A practice built on the first sentence will still be standing after the second is forgotten.