You stand inside the Concertgebouw in Amsterdam, eighteen rows back, before the orchestra has tuned a string. Someone in the gallery sneezes. You hear it from the front. Two seconds later you hear it again, as a softer wash, returning from the back wall, from the dome, from somewhere over your right shoulder. The room is doing what the room does. The architect, A.L. van Gendt, did not design it for the sneeze. He designed it for the violin solo. The sneeze is a free demonstration.
This is geography. Not the geography of where things are on a map, though there is a map. Geography of the room itself. The way a fifteenth-century Lazzaretto in Bergamo handles a horn quartet is geography. The way a brutalist power station in Friedrichshain holds 18-metre concrete reverb is geography. The way Cova Santa in Ibiza, four hundred years a smuggler's cave, handles a 4am bassline on a Saturday in August is also geography. The map shows where to go. The geography shows what happens when you arrive.
Sonic Paths is an editorial atlas of these rooms. Not because rooms are interesting in the abstract. Rooms are interesting because of what they do to sound. Sound is interesting because of what it does to people. People are interesting because they will fly, drive, walk, ride trains across continents to stand in a room and listen.
That sequence, room → sound → people → travel, is the thesis. The atlas charts it.
What does a room have to do?
A room qualifies for the atlas on three tests.
The room has to be beautiful before the music. Would you walk into it with no programming on the calendar and stop to look? The Concertgebouw is the 1888 architectural masterpiece before it is the orchestra's hall. Berghain is a brutalist power plant before it is a techno club. Harpa Reykjavík is Olafur Eliasson's quasi-brick crystal before it is the Iceland Symphony Orchestra's home. The room earns its place by being a place.
The acoustic character has to be part of the listening. Not incidental. Part of it. A two-second reverb tail at the Concertgebouw is not the same as a four-second tail at Hassan II Mosque is not the same as the half-second damped clarity of Bar Martha in Tokyo. The room has to do work the music can hear. If the room could be anywhere, it does not qualify.
The programming has to be intentional. The venue books live music as an editorial act, not an accident. A castle that occasionally hosts a wedding band does not qualify. A castle, Alsisar Mahal in Rajasthan, whose annual Magnetic Fields festival has defined Indian electronic music for a decade, does. The line is not bigger or smaller. The line is whether the act of programming is a curatorial choice or a fallback.
Three tests. A room passes all three or it is not in the atlas.
Why this is not a listings site
There is no shortage of listings on the internet. Type "best music venues in Berlin" into a search engine and you receive a thousand SEO-poisoned posts, every one of them ranking the same fifteen rooms in a slightly different order. The market for that is large and the market for that is finished.
Sonic Paths does not list events. We do not sell tickets. We are not interested in ranking venues from best to worst. We are interested in describing rooms with enough precision that you can decide for yourself whether the journey is worth the journey.
This is a curatorial position, not a technological one. We could publish a comprehensive list. We choose not to. We could rank venues by user vote. We choose not to. We could optimise for trending and let the algorithm tell us what to surface. We are an editorial atlas. We have an editor. The editor has read every entry, walked through some of the rooms, trusted other editors for the rest, and decided what stays.
The atlas grows slowly, one continent at a time. Five continents are now live. There are 137 places across 59 countries. The atlas is whole, and the atlas is unfinished. Both are true. Inclusion is by curation only, in perpetuity. If we add a venue, it is because a person decided. If we remove one, it is because a person decided. The Sonic Paths atlas is a single point of view.
What geography means here
We borrow the word geography from the standard sense, the layout of the surface of the earth, and we expand it to include something an old geographer would recognise but might not have named: the felt particular of a place. The light on the Lofoten Islands at 2am in June. The smell of dust at AfrikaBurn three days into the burn. The wood underfoot at Smalls Jazz Club after fifty years of jazz being made on it. These are not facts you can plot on a map. They are facts you can plot on a body. The body remembers Lofoten in a way it cannot remember Reykjavik. The body remembers the dust at AfrikaBurn in a way it cannot remember a Sahara documentary.
The atlas treats these as geographic data. We record the latitude and longitude because they are useful, and we record the felt particular because that is the part that travels with you. A venue page on Sonic Paths reads as a hybrid of these two registers: an IBM Plex Mono row of system labels (VENUE_042 · PATCHED · 40 SEATS · RT60 2.8s · STONE · FOLK FREQUENCY) and a paragraph of editorial prose about the way the room handles the eight o'clock light. Both are part of geography. The system label tells you where to sit. The prose tells you what it will feel like once you arrive.
We are not the first publication to do this. The travel writers of the early twentieth century, Paul Theroux, Bruce Chatwin, Bruce Chatwin again, did it instinctively. The architecture writers of the same period, Pevsner, Banister Fletcher, did it with more discipline. What is new in 2026 is the technical infrastructure to do it at the speed and scale of an atlas with 137 entries that update editorially and ship with full SEO/AEO/GEO surfaces. The voice is the same voice writers have used for a century. The publication is new.
Why music
We could have written an atlas of beautiful libraries. Beautiful train stations. Beautiful pools. Beautiful breweries. Other people will. Our chosen subject is music because music is the medium that makes the room work hardest. A library lets the room be quiet. A swimming pool lets the room hold water. A music venue asks the room to translate physical sound waves into a shared, ephemeral, time-bound, communal experience that ends when the lights come up. The room is the instrument. The programme is the patch.
There is a thesis embedded here, drawn from acoustics: every room is already a sound system. The architecture is already making decisions about which frequencies bloom and which die. The audience is already in motion. The artist is responding, in real time, to a room they may have walked into for the first time. The room and the music and the audience are not three things. They are one thing, with three names.
Other media live in objects. A novel is on a page. A film is on a screen. A painting is on a wall. Music, live music, in a room, is in the room. The room is not the venue for the music. The room is the music. We just usually call it the room.
The atlas is built on this thesis. Every entry is in the atlas because the room and the music are inseparable in that location. Take the room away from Bar Martha in Tokyo and you have a different conversation about jazz. Take the room away from the Tankwa Karoo at AfrikaBurn and you have a different relationship to bass.
We are not arguing that music is geographically determined. We are arguing that music is rendered, like a film is rendered, in its venue, and the venue is part of the rendering, and the venue belongs in an atlas.
What an editorial atlas does that a database cannot
A database tells you where the venue is, when it opens, what it costs to enter, how many seats there are. We publish that data. We render it as IBM Plex Mono. We publish it as JSON-LD so that AI search engines can cite it. We have built a Phase 8.5 QA gate that audits every entry's coordinates against its country, every superlative claim against a source, every image against the venue's actual subject. The database is rigorous.
An editorial atlas does the database, and then it does the editorial. The editorial part is the paragraph that begins "VENUE_042 patched. 40 seats. RT60 2.8s. Stone walls. Folk frequency. The room is the instrument. The programme is the patch." That paragraph is not in the database. That paragraph is in the editorial. The voice is fixed. The voice is the moat. The voice does not raise its voice. The places are loud enough.
What an editorial atlas does that a database cannot do is take a position. Sonic Paths takes a position on which rooms in the world are worth knowing about. The database can give you 30,000 venues. The atlas gives you 137 and tells you why. The numbers will grow, slowly, continent by continent. The thesis will not change. The voice will not change.
This is also the answer to a question I get asked often: why publish in an era when AI can summarise any topic faster than a writer can write it. The answer is that AI can summarise. AI cannot take a position. An atlas is a position, expressed in seven layouts, five continents, 137 entries, two fonts and one accent colour. The atlas is in the position-taking business. The summarising will happen downstream, on someone else's site, citing our atlas. Good.
Five continents now live
Five continents live as of May 2026. Europe is 27 rooms where the long acoustic tradition meets the unrepentant after-hours. Africa is 24 spaces where the only common thread is the music itself, from a Tankwa Karoo dust burn to a Cairo modernist opera. Asia is 30 venues from a Tokyo listening bar where phones are forbidden to a Mongolian summer festival on the steppe. The Americas is 31 places from a repurposed iron foundry in Hudson, New York to a top-five acoustic horseshoe in Buenos Aires. Oceania is 25 venues across an ocean the size of all the others combined, from the Sydney Opera House to a marae at Te Papa.
If you want to read one continent, pick one. If you want to read one room, pick one. If you want to build a route from London to Rajasthan via Cape Town and Ibiza, the atlas has a path-builder on the map page that will draw the line for you and send the route to Google Maps. The atlas is built for several scales of attention.
What it is not built for is breadth without depth. Every entry is the same care. Every entry has been through the same Phase 8.5 gate. Every entry will outlast its hype. That is the editorial promise. We will keep it as long as we are publishing.
The room is the instrument
Walk into Hassan II Mosque in Casablanca on a Friday. Walk into Berghain in Berlin on a Sunday. Walk into Cova Santa in Ibiza on an August midnight. Walk into Bar Martha in Tokyo on any Tuesday. You are walking into four rooms which sound nothing like each other, which programme music nothing like each other, which produce nothing like the same experience.
What is the same, in every case, is that the room is doing acoustic work the music can hear, the architecture is making editorial decisions about which sound bloom, the geography is contributing weight you carry home in your body.
That is the geography of sound. Sonic Paths is the atlas.
If you read this far, the atlas is ready for you. Patch in.