TyfloPodcast
A podcast platform about technology and accessibility for blind and partially sighted people. I am its founder and editor-in-chief.
I don't separate work from passion. It's one organism. Below are the things I create, co-create and maintain.
A podcast platform about technology and accessibility for blind and partially sighted people. I am its founder and editor-in-chief.
An internet radio station and its main channel. I am a co-creator of the channel format.
A tool for auditing digital accessibility. It analyses websites and documents against standards (WCAG, PDF/UA). Beta version.
Simple Accessible Radio Automation, an accessible radio automation system. Open-source software written in Python.
TyfloPodcast clients for three platforms: Windows, Android and iOS. The interface is designed for accessibility and screen readers (NVDA, TalkBack, VoiceOver).
An accessible frontend for e-podroznik.pl: a connection search and stop timetable, friendly to screen readers.
An app for viewing TV and radio programme schedules, friendly to screen reader users.
A set of accessible Windows GUI apps for music recognition (Shazam): from a file, from audio input or output, and updating “now playing” on Shoutcast and Icecast streams.
An accessible Windows GUI browser for Usenet newsgroup archives (Usenet Archive Toolkit).
A website for Monika Machul’s craft studio “W Deseń”. I built it with the Astro framework and additionally created a convenient tool that lets the author publish photos of her work herself — adding them, organising them into categories and pushing them live without touching any code.
An NVDA driver for the Dolphin Apollo 2 hardware speech synthesizer (and Juno). It lets you use these synthesizers as the voice of the NVDA screen reader. Distributed as an NVDA add-on.
A simple, accessible (NVDA/JAWS) Windows app for removing all digital signatures from PDF files.
A Windows app that automatically removes filler sounds like “um” and “uh” from speech recordings and shortens overly long pauses. It detects them locally, on the user’s own computer, with the classla/wav2vecbert2-filledPause AI model (nothing is sent to the cloud). It sounds natural, is fully accessible to screen readers, and can export a ready-made Reaper project.
A laboratory AAC audio encoder for power users. It takes the Fraunhofer library (libfdk-aac) and exposes its normally hardcoded internal decisions as command-line switches: per-band joint stereo control, SBR density, parametric stereo, noise shaping, core bandwidth cutoff and much more. A tool for experimenting with and inspecting what the encoder does on its own. Subject-matter consultation: Patryk Faliszewski.
DSP plugins for listening to audiobooks, podcasts and spoken-word recordings: Bookamp for Winamp, Bookbar for foobar2000. They smoothly change tempo without changing pitch, pitch without changing tempo, and a „tape” mode that moves both at once. They have two time-stretch engines, optional loudness leveling and speech intelligibility enhancement, plus keyboard shortcuts. The interface is bilingual and fully accessible to screen readers. Subject-matter consultation: Patryk Faliszewski.
QAT (Quality Audio Tester) is a minimalist DSP plugin with no window and no settings, controlled with two keys — available for Winamp and for foobar2000. It is meant mainly for checking the mono compatibility of a recording: the Q key switches between stereo, the mono sum and the difference signal, and the W key toggles speech clarity. Subject-matter consultation: Patryk Faliszewski.