Телеграм канал «🇮🇳 Android Knowledge for the Blind 🧑‍🦯 📱»

Kokoro TTS: Private, Offline English Voices for Long-Form Reading

Enjoy natural-sounding Kokoro voices in your Android reading apps—privately and offline.

SC Kokoro is a system text-to-speech engine designed for books, articles, and other long-form content. On compatible phones, it uses the GPU to handle speech generation. Compared with CPU-only Kokoro engines, this can help capable devices prepare upcoming speech and make reading feel more responsive. It can also reduce heat and battery use, depending on your device and reading setup.

Choose from 28 English voices, including US and UK options. Download the voice data once, then generate speech on your device. The text you listen to is not sent to a cloud speech service.

SC Kokoro works with apps that support Android text-to-speech engines. Select it in a reading app that lets you choose a TTS engine. For smooth long-form listening, we recommend Speech Central, whose advanced buffering can prepare upcoming passages and help reduce interruptions. Playback in other apps may vary depending on how they request and buffer speech.

Speech generation speed depends on your device. With effective buffering, long-form reading can work well even on mid-range devices. On slower devices, SC Kokoro can still generate high-quality audio files in apps that support audio export, though the process may take longer. Tasks that require immediate spoken feedback, such as TalkBack, are more demanding and may be challenging even on high-end devices.

SC Kokoro checks compatibility before downloading the model and provides an estimate of the experience you can expect. For best performance, keep the screen off while listening.

Source code and licenses: https://github.com/speech-central/sc-voices-android
Download: https://play.google.com/store/apps/details?id=com.labsii.voices

🌟 Introducing Sharada Braille Writer (SBW) for Windows! ⌨️⠿

Sharada Braille Writer (SBW) is an advanced text editor designed to support Braille input. It allows simultaneous multiple key presses of Braille combinations to produce corresponding letters seamlessly. Keys f, d, s, j, k, and l represent Braille dots 1, 2, 3, 4, 5, and 6, respectively, enabling an intuitive typing experience for Braille users.

✨ Key Features:

🔹 Six-Key Braille Input: Type Braille combinations using a standard keyboard.

🔹 Multilingual Support: Supports more than 50 languages through Liblouis, along with over 30 built-in language options.

🔹 AI Integration: AI integration to enhance the user experience.

🔹 Accessible Learning: Helps visually impaired students practise Braille typing and participate more independently in digital education.

🔹 Free Trial Available: Try SBW before purchasing a licence!

🔹 Windows Support: Available through the Microsoft Store.


💰 Special Pricing for Multiple Licences

Schools, institutions and organisations can benefit from reduced pricing when purchasing multiple licences:

✅ 5 licences: ₹1,500 per licence
✅ 10 licences: ₹1,000 per licence

🔗 Download Sharada Braille Writer and try it today:
https://apps.microsoft.com/detail/9pmlt82z3zfq

🤝 Interested in purchasing multiple licences? Contact Zendalona for more information about institutional pricing.

💙 Developed by Zendalona — Building an inclusive digital world through accessible technology!

📢 Please share this with visually impaired students, teachers, schools and organisations that may benefit from Sharada Braille Writer.

#SharadaBrailleWriter #Braille #Accessibility #Windows #Zendalona #InclusiveEducation

#latest update: What's new? Downloaded filenames now include the uploader and YouTube video ID, so different videos with the same title do not overwrite or skip one another.

Note! Run it directly while your current A11YTube installation is present; you do not need to reinstall the old installer first. Setup looks for the existing install location and updates its program files in place. If it can't detect an older portable folder, select that folder on the destination page. Please close A11YTube before updating.
The app keeps its settings in
%APPDATA%\ddt.one\A11YTube\settings.ini
and its history, favorites, collections, and continue-watching positions in
%APPDATA%\ddt.one\A11YTube\A11YTube.db
. The installer and uninstaller leave that data folder untouched.
The settings and database remained unchanged. If you want to send your feedback, contact me on my Telegram account telemetr.me/Djgsdjgsd. Updated by darry john sarmiento! Special credits to @Daoductrung

Check out "Accessimind Screen Reader"
https://play.google.com/store/apps/details?id=com.sarperarikan.accessimindreader

Translations and documentation have been updated to reflect the version 1.0 release. We have added full interface and documentation support for Spanish and Brazilian Portuguese, and the Simplified Chinese interface and documentation have been completely retranslated to align with official NVDA terminology.

Google TTS for NVDA 1.0 Release: Enhanced Performance, Stability, and Reliability

Developers: Nguyễn Anh Đức, Đào Đức Trung, and Phạm Hùng Vương.

We are pleased to announce the release of Google TTS for NVDA version 1.0, representing a significant milestone in the project's maturity. With a supported Chromium browser runtime now integrated, Google TTS serves as a highly dependable solution for daily NVDA usage. This update facilitates faster speech initiation, smoother playback, improved multilingual text processing, and more accurate automatic language profile selection. Furthermore, it provides consistent audio volume and robust recovery mechanisms in the event of errors within the Chromium runtime, the WASM TTS Engine, or NVDA’s audio output.

### Key Highlights of Version 1.0

1. Optimized Reading for Medium and Long Text
Google TTS now initiates speech for medium and long passages more rapidly by processing the initial segment immediately, while handling subsequent text in larger, more efficient blocks.

Once audio is generated, the add-on caches completed segments in RAM, allowing for instantaneous reuse when identical content or speech settings are requested. For initial playback of long passages, the engine buffers a small amount of audio to ensure seamless continuity. While responsiveness remains subject to the inherent behavior of the original engine and the availability of cached data, these optimizations significantly enhance the overall user experience.

2. Enhanced Multilingual Text Processing
Sentence and phrase segmentation now leverage the comprehensive Unicode 17.0 and CLDR 48.2 standards. This upgrade improves the handling of localized punctuation, abbreviations, URLs, emojis, and non-space-delimited scripts, including Chinese, Japanese, Korean, and Thai.

The "Automatic Language Profile" feature utilizes this expanded linguistic data to select the most appropriate profile with greater precision, even when the input text is brief or ambiguous. Additionally, pauses and transitions between segments are now more consistent, regardless of whether the audio is retrieved from the cache or newly generated.

3. Improved Volume Consistency and Clarity
Version 1.0 implements fixed amplification to increase default loudness while eliminating the erratic volume fluctuations observed in previous iterations. The volume setting remains proportional, and protective measures against excessive audio peaks have been maintained.

*Note:* Some male voices may exhibit residual distortion; this is inherent to the original WASM TTS Engine output. This release focuses on optimizing loudness and consistency rather than modifying the underlying engine-side audio characteristics.

4. Robust Recovery and Reduced Conflicts
The add-on now features an improved recovery protocol: if the Chromium runtime or WASM TTS Engine fails prior to audio generation, Google TTS will discard the compromised instance and attempt a single retry with a fresh runtime. To prevent redundant or fragmented speech, retries are suppressed once audio playback has commenced.

The "Keep Google TTS browser runtime ready" feature has been refined to ensure that busy, disconnected, or unhealthy runtimes are not retained. Furthermore, the add-on now exclusively loads its own bundled components, effectively mitigating conflicts with similarly named files from other NVDA add-ons. Recovery is also more resilient during changes to NVDA’s audio output device or when encountering system-level audio errors.

5. Updated WASM TTS Engine and Expanded Localization
The bundled WASM TTS Engine has been updated to version 20260917.1, and the voice catalog has been refreshed to reflect Google’s latest revisions.

Important: If you have previously installed voice packages, please open the Voice Manager and re-download (overwrite) them. Legacy voice packages are incompatible with the new version, as their file structures must align with the updated integrity verification standards.