sound flow · city upstream
一条由声音构筑的流动场域。
留下你的呼吸，看它如何随时间被河水慢慢遗忘。
A field built of voices. Leave a breath; watch it be slowly forgotten by the river.
触碰，进入场域 touch to enter场域分为三层并行的声音生态——它们彼此叠映，永远不会同时清晰。 The field is three overlapping sound ecologies. They never resolve into one.
访客投入的声音。在河流上自西向东缓慢漂流， 沿途经历五层熵蚀算法的风化，最终在入海口被服务器物理删除。
Voices left by visitors. They drift east, weathered by five layers of entropic decay, then erased at the estuary.
固定在河流特定坐标上的声源——城市录音、潮汐基底、人为植入的场域底色。 它们不漂流，但会随时间日夜呼吸。
Sources pinned to specific river coordinates — city recordings, tidal bass, curatorial undertones. They breathe with day and night, but do not drift.
WebGL 流体模拟与地图叠映。流速、河段、地形、潮汐—— 所有声音的水文学都被同步绘制为可见的形态。
WebGL fluid simulation overlaid on the map. Every sound's hydrology — flow, zone, terrain, tide — is rendered as a visible shape.
没有目录，没有搜索，没有重播。只有此刻。 / no index, no search, no replay — only now.
我们不在建造产品。我们在养一条河。
声音 被河水 慢慢风化—— 熵从听觉的边缘渗入，颗粒化、相移、比特蚀， 最终在入海口被服务器 物理删除。
所有访客共享同一条河。 你听到的，取决于你站在哪里； 你留下的，没人会再听到原样。
we are not building a product. we are tending a river. sounds are weathered downstream — granular, phase-shifted, bit-eroded — and finally erased from the server at the estuary. every visitor shares one river; what you hear depends on where you stand, and what you leave is heard by no one in its original form.
每段河流对应一种声音处理。你投入的位置，决定了这段录音将被如何重塑——从几乎完整，到彻底消融。
Where you drop your voice along the river decides how it's treated — from nearly intact to fully dissolved.
场域无目录、无搜索、无重播。下面是三个真实的状态截屏——
入场、留声、聆听。
The field has no index, no search, no replay. Three states: enter, record, listen.
