Editorial standards
How we research
Where the information on this site comes from, what we will not claim, and how the music itself is made.
37
tradition monographs
29
instrument & gear guides
166
picks, each spec-sourced
261
listings verified
What we actually read
Manufacturer specification pages, linked per pick
Every gear recommendation is built from the maker's own published specification — body timber, scale length, pickup configuration, sample rate, driver size, impedance. All 166 of the 166 picks across our guides carry the URL of the spec page it came from, so any figure we print can be checked against its source rather than taken on trust. Where a manufacturer does not publish a figure, we say it is not stated instead of estimating it.
Owner reports, read for patterns rather than quoted
We read owner reviews and player forums for what recurs — the neck that goes sharp in dry air, the interface driver that fights a particular laptop, the pad that flattens by month eight. We look for the failure mode several hundred people describe, not the one furious paragraph. We do not reproduce individual posts, usernames or handles, and we never invent one.
Published scholarship, for the traditions
The monographs draw on published histories, instrument organology, museum and archive documentation, and the writing of musicians inside each tradition. Where a date, an origin or an attribution is contested — and in this subject it very often is — the page says it is contested rather than picking the tidiest version. Where a claim traces to a specific named writer or source, we name them in the sentence.
Listening
Recordings of the instruments and traditions themselves, including performances by practitioners who learned them from inside. That is the one part of this we really do first-hand, and it informs the descriptive writing rather than any factual claim.
The SlowHum Score
What the number on a pick means
Some picks in our guides carry a SlowHum Score between 62 and 98. It is arithmetic, not a verdict: two published figures combined by a fixed formula, recomputed whenever the underlying data changes. It is not a test result. Nobody here played, measured or recorded with any of these instruments for this site, and no part of this score comes from doing so.
Score = 0.72 × owner satisfaction + 0.28 × value
Rounded to a whole number and clamped to 62–98.
Owner satisfaction — 72%
Amazon’s own star rating for that exact listing, plus a small confidence bonus that grows with the number of ratings behind it, so a 4.7 from eleven thousand owners counts for more than a 4.7 from ninety. Where Amazon also publishes its programmatic review-aspect tallies — how many customers mentioned build quality, value, sound, and whether they did so positively — the positive share of those tallies is blended in, but only when at least fifty qualifying mentions exist. This is other people’s experience of owning the thing, aggregated. It is the closest honest substitute we have for hands-on use, and it is deliberately the dominant term.
Value — 28%
Price position inside the product’s own guide — the set of options a reader is genuinely choosing between on that page, not the whole catalogue. The cheapest pick in a guide scores 95 and the most expensive 60, straight-line between. Where we have no live price, or a guide’s picks are all one price, this term takes a neutral 78 rather than a position we cannot justify.
Most of the catalogue has no score, on purpose
Of the 261 listings behind our guides, 99 are scored — 38% of them. 151 carry no published Amazon rating at all, and a further 11 are rated by fewer than twenty-five customers, which is not a crowd. Both groups show no score. Owner satisfaction is nearly three-quarters of the formula: filling it with a default, or letting five stars from three people top the table, would mean printing a number we did not have and calling it a measurement. A missing badge means unrated or thinly rated, not poor — plenty of unscored picks are the best answer on their page, which is why they are on it.
What it cannot tell you
It cannot tell you whether an instrument suits your hands, your room or your ear, and it does not know that a model has been superseded by a newer one. Two picks a point apart are not meaningfully different. The written verdict on each pick is the part that is actually reasoning about your situation; the score is a summary of what everyone else thinks and what it costs.
What we will not do
We do not test instruments or gear
There is no SlowHum test bench, listening lab or review inventory. Nobody here played these guitars, tracked through these interferences or wore these headphones for a week. You will not read “we tested”, “hands-on” or “we recorded with this one”, because none of it would be true. You will read “researched” and “compared”. Where genuine hands-on testing exists in the world, we weight it and say whose it is.
We do not claim expertise we do not have
We are not ethnomusicologists, instrument makers or teachers of any of these traditions. The monographs are careful, sourced, respectful reading — not scholarship, and not a substitute for learning from someone inside the tradition. Where a living practice has custodians, they know more than this website does, and the pages say so.
We do not invent a person, a quote or a figure
Work here is published under the brand, not under a fabricated byline or an invented reviewer. No composed quotations from players or scholars, no numbers that sound precise but trace to nothing. A statistic we cannot follow back to a primary source does not appear at all.
We do not hide how the music is made
The pieces on the SlowHum channel are produced with the assistance of modern music-generation tools, guided by research into the tradition each draws from. Every video carries YouTube's Altered Content disclosure. This is stated plainly here and on the about page because a listener deserves to know what they are hearing, and because a site that hid it would have no standing to write anything else on this page.
We do not take payment for placement
No maker or retailer pays to be recommended, and we accept no sponsored placements or free instruments. The guides carry affiliate links, disclosed on every page that has one; the commission does not change the order of the picks or which ones appear at all.
We do not print a freshness date we did not earn
A guide or a monograph shows the date it was published. It shows an “Updated” date only when the writing itself was actually revised — not because a price refreshed or a field was regenerated. Most pages here therefore show a publish date and nothing more, which is the honest answer. The same date goes into the page's structured data, so a crawler is told exactly what you are shown.
If something here is wrong, it should be corrected rather than defended. A guide changes when a maker publishes a spec we did not have, when a recurring fault shows up in owner reports that was not visible when we wrote it, or when a model is discontinued and the honest answer is a different one. A monograph changes when better sourcing contradicts it.
More about the project and its disclosures is on the about page.