FlashRead: RSVP & High-Speed TTS Engine

FlashRead (fork of Voxki Custom) transforms Anki into a high-speed reading tool. It combines RSVP (Rapid Serial Visual Presentation) engine with a lightning-fast offline text-to-speech system, designed to drastically cut down review times while maximizing reading comprehension. Unlike other TTS addons, this one emphasizes audio pre-generation to minimize lag and optional saved audio for use on ios/android devices.

🚀 High-Speed RSVP Reading Engine

Read flashcards at up to 1,500 WPM without moving your eyes. The RSVP engine flashes words sequentially in the center of your screen, eliminating the time your eyes waste scanning across sentences.

• Cognitive Science ORP Tracking: Automatically highlights the Optimal Recognition Point (ORP) of each word in red to lock your visual focus.
• Advanced Dynamic Pacing: The engine naturally mimics human reading rhythms. It automatically slows down for long words, pauses dynamically at commas and periods, and gives you extra time to process formatted text (bold, italics, underline).
• Distraction-Free Themes: Choose between Light, Dark (Translucent), and Dark (Pitch Black) overlays to block out the rest of the card while reading.

🔊 Lightning-Fast Native Text-To-Speech

FlashRead features a zero-latency, fully offline text-to-speech engine built directly on top of your operating system's native architecture. No cloud APIs, no internet connection required.

• Smart Text Filtering: Automatically strip out unwanted text. Easily configure the engine to skip text inside (parentheses), [brackets], or {braces}.
• Custom Pronunciation Dictionary: Sick of your TTS engine mispronouncing medical acronyms? The built-in dictionary lets you effortlessly map raw text to phonetic sounds (e.g., automatically read "CHF" as "C H F" or "CABG" as "cabbage").

🍎 macOS vs 🪟 Windows TTS Differences:

While the RSVP visual engine is fully cross-platform, the TTS engine relies on OS-specific architectures:

• macOS (Highly Optimized): Uses Anki's native say engine with an advanced Pre-Generation Pipeline. FlashRead predicts your upcoming cards and pre-caches the audio to your hard drive before you even flip the card, resulting in absolute zero-latency playback. Also supports assigning different voices (e.g., Daniel vs. Karen) to the Question and Answer sides.
• Windows: Automatically detects your OS and falls back to Windows' native SAPI (System.Speech.Synthesis) via PowerShell. Audio is streamed live rather than pre-generated, which may result in a tiny fractional delay before playback begins, but is otherwise mostly functional.

⚙️ Controls & Customization

FlashRead is built to be tweaked to your exact preferences. Access the settings via the Anki top menu bar (FlashRead > Settings...).

• Global Hotkeys: Easily adjust your WPM or audio speed on the fly directly from the reviewer without opening the settings menu.
• Targeted Activation: Choose exactly what side of the card FlashRead runs on, or restrict it to only read specific extra fields.