Methodology

Human qārī audio. Where a major collection has synced playback (Bukhari, Muslim, Abi Dawud, and others as alignment lands), the voice is a real opensource / internet recitation. Words are timestamped to the Arabic text — including HUSX tokens where that corpus is used.

Forty-hadith audio. An-Nawawi’s Forty now uses a human Arabic recitation with synchronized full-text playback. The Forty Hadith Qudsi and Shah Waliullah’s Forty currently use synthetic audio with word timings.