the geometry of shibuya-kei: a shelf of neighbours for long sessions

1. the shelf has no label on the spine

i do not keep a playlist called shibuya-kei. cornelius and pizzicato five are the names people give as the front door to the genre, and i am writing from the neighbours' side of the street. the queues i actually keep have long, specific names:

• overengineered pop: sophisti-pop, city pop, chamber pop, yacht rock and boogie in one queue. it runs to 35 tracks.
• 21.0 fm made in japan: the japanese side of the shelf, 60 tracks.
• sleepy japanese jazz: 35 tracks for when the room should go quiet.

the same account also plays a lot of rap and guitar bands. this is one shelf, not the whole house.

2. records you reach for after years

a record you still reach for after years works like a jacket you still wear after years. the wear is the proof. hikaru utada is the top artist on my long-term list, with simple and clean and movin' on without you among the tracks, and lamp sits in the same list. kirinji's kuruma to onna shows up in my top tracks for the last four weeks, the last six months and the long run, the last time as its 2018 remaster. that is a reference track. you know what the room should sound like because you have heard it there before.

3. the bar behind the record

the description on sleepy japanese jazz says the lupin bar in yokohama is my dream destination when i'm older. that sentence explains the playlist. a bar with one counter, low light and one record at a time is a small room with a fixed tempo. the playlist tries to build that room around a desk. it is quiet on purpose, and nothing in it wants an answer.

4. the same warmth, closer to home

the shelf has an indonesian wing, and it is my longest playlist. 80's indonesia city pop/pop kreatif/jazz fusion runs to 348 tracks. chrisye tops my six-month artists, with marlina among the tracks. white shoes & the couples company and diskoria are in the current rotation. the track at the top of all three time windows is kafin sulthan's hukum murphy, and murphy's law is the same idea i build with: expect failure, map the edge cases early, reka peluang.