Librezam od FoxRefire

This is a browser extension that uses internal API of Shazam to recognize the music playing in the current tab.
Librezam also supports Audd and ACRCloud as alternative backends, which can be changed from the configuration.
Unlike other similar extensions, it works on almost all sites and has fewer bugs.

Requests unrelated to music recognition, such as telemetry, will not be made.
This is unofficial project and is not affiliated or supported by Shazam ltd.

!!Please report extension issues to Github Issue or ‘bqzfjohg@addy.io’, not in the review section, and read the troubleshooting below before reporting the problem !!

Source code
https://github.com/FoxRefire/Librezam

Comparison with similar extensions

• Song Identifier
  Backend supports ACRCloud only (does not support humming response)
  No history, auto-recognition mode, auto-retry, and microphone recognition feature
  Does not support CORS, Headless Element, and ShadowRoot detection
  Open source (Unlicense)
  API key registration required (up to 100 requests per day free of charge)
  Telemetry unrelated to music recognition will not be performed.
• Audd
  Backend supports Audd only
  No auto-recognition mode, auto-retry, or microphone recognition feature
  Source-available proprietary
  Supports CORS, Headless Element, and ShadowRoot detection starting from version 3.2.4 (Fun fact: I was the one who advised on how to fix these issues with the Audd extension)
  Some sites like Twitch display a message prompting you to purchase an API key
  Not only audio data, but also telemetry such as tab URLs are transmitted.
• Librezam
  History functionality
  History CSV export
  Supports multiple backends (Shazam, Audd, ACRCloud)
  Shazam backend can be used without API registration and without rate limits(ACRCloud backend requires API key settings)
  Humming and cover song recognition (ACRCloud backend)
  Automatic retry with larger recording length and another backend when recognition fails
  Automatic recognition mode
  Microphone recognition mode
  Supports detection of CORS, Headless Element, and ShadowRoot (also plans to support detection of AudioBufferSourceNode)
  Open source (AGPL-3.0)
  Cross-browser compatibility (also works on Firefox for Android and Chromium)
  Telemetry unrelated to music recognition will not be performed.

Comparison of each backends

• Shazam
  No registration required, no rate limit
  Privacy-conscious design that sends recorded audio to the server as a fingerprint that cannot be traced back to the original content, rather than sending the audio itself
  Recognition speed is slightly slow
  High recognition rate
• Audd
  No API key registration required for up to 10 requests per day
  Moderate recognition rate
  Recorded audio is sent to the server as is
• ACRCloud
  API key registration is required, up to 100 requests per day are free
  Can recognise humming and cover songs
  Very high recognition rate
  Recorded audio is sent to the server as is
• Tencent
  No registration required, no rate limit
  High recognition rate
  It has a very high recognition rate, especially for anime/game music and Chinese music. For example, this song cannot be recognized by any music recognition service, including Shazam, Audd, ACRCloud, or Google, but only by the Tencent backend.
  https://www.youtube.com/watch?v=ktOpWwiTFMI
  High noise and data loss tolerance
  Recorded audio is sent to the server as is