Data Provenance, Streaming Methodology & Catalog Telemetry

ice Radio FM maintains an open, rigorous, and automated directory of global internet radio stations. This document outlines our data provenance, stream health verification protocols, deduplication pipeline, and directory quality standards.

1. Stream Ingestion & Provenance

Station metadata is aggregated from multiple authoritative sources:

• Direct Broadcaster Submissions: Broadcasters register and verify their stream URLs, logos, codecs, and schedules directly via our Broadcaster Portal.
• Public Directory Feeds: We cross-reference public radio directory registries (including Radio-Browser.info and open SHOUTcast/Icecast yellow pages) under open licenses.
• Automated ICY Header Scrapers: Streams are periodically polled to parse icy-name, icy-genre, icy-br (bitrate), and icy-metaint headers directly from origin streaming servers.

2. Stream Health & Telemetry Verification

Unlike static link directories, ice Radio FM actively monitors stream health. Automated background agents perform round-the-clock telemetry checks:

• HTTP & Audio Handshake: Validates HTTP status codes (200 OK), audio MIME types (audio/mpeg, audio/aacp, audio/ogg, application/vnd.apple.mpegurl), and network latency.
• HLS & Chunk Inspection: For HLS (.m3u8) streams, our parser verifies that media playlists and audio segments are accessible and actively updating.
• Uptime Scored Retention: Stations undergoing persistent outages are automatically flagged and suppressed from search recommendations to ensure listeners never hit broken audio.

3. Supported Audio Codecs

4. Deduplication & HTTP 410 Tombstoning

To protect search engine index hygiene and prevent soft-404 dead ends, ice Radio FM implements strict URL canonicalization:

• 301 Permanent Aliasing: Station slug variants or renamed broadcasters are permanently redirected to their canonical station page.
• HTTP 410 Gone Tombstones: When a station permanently ceases broadcasting, our database registers a StationTombstone. Requests to these URLs immediately return HTTP 410 Gone, prompting search engines to quickly and cleanly deindex the page without wasting crawl budget.
• Soft-404 Elimination: Non-existent slugs return HTTP 404 rather than redirecting to search results, adhering strictly to Google and Bing webmaster guidelines.

5. Open AI & Machine Retrieval (GEO / AEO)

We support open data discovery for AI search engines (ChatGPT Search, Perplexity, Google AI Overviews, ClaudeBot):

• llms.txt: Structured index of authoritative endpoints and API capabilities formatted for large language model retrieval.
• Schema.org Microdata: Full RadioStation, CollectionPage, BreadcrumbList, and FAQPage JSON-LD structures embedded in server-side rendered HTML.
• Clean Internationalization: Zero duplicate query params; reciprocal hreflang links across all supported language prefixes.