<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Tracks on TinyChen's Blog</title><link>https://tinychen.com/tags/tracks/</link><description>Recent content in Tracks on TinyChen's Blog</description><generator>Hugo</generator><language>zh-cn</language><lastBuildDate>Thu, 23 Jul 2026 12:00:00 +0800</lastBuildDate><atom:link href="https://tinychen.com/tags/tracks/index.xml" rel="self" type="application/rss+xml"/><item><title>Octans-媒体库轨识别与动态范围探测</title><link>https://tinychen.com/20260426-octans-media-track-dynamic-range/</link><pubDate>Sun, 26 Apr 2026 12:00:00 +0800</pubDate><guid>https://tinychen.com/20260426-octans-media-track-dynamic-range/</guid><description>&lt;p&gt;本文主要说明自托管媒体库（以 Octans 扫描与轨事实层为例）如何把「容器内 stream 事实」和「给 UI / NFO / 选轨用的展示轨」分层建模，如何用语言与标题启发式生成可读轨名，以及视频动态范围（HDR / Dolby Vision 等）为何要按证据优先级保守分类、并在必要时做分层探针。探测工具选型本身见同系列 MediaInfo / ffprobe 文，本文不重复速度与字段面对比。&lt;/p&gt;</description></item></channel></rss>