ClaroRead

ClaroRead covers similar ground to other literacy toolbars: text to speech with highlighting, study tools and writing support. What sets it apart in practice is how well it handles material that was never made accessible in the first place.

A scanned PDF, a photograph of a page, a form that arrived as an image. These are the documents that stop people, and turning them into readable text is where ClaroRead earns its place.

• Screenshot reading is the headline feature. Capturing an area of the screen and having the text recognised gets you into documents that no amount of arrow keys will open.
• Scan quality decides everything. A straight, well lit page at a decent resolution recognises almost perfectly. A phone photograph taken at an angle does not.
• Check columns and tables afterwards. Recognition reads a page in the order it thinks the text runs. Newspaper columns and complex tables are where it most often gets that wrong.
• Save the recognised text. Once you have a clean version, keep it. Recognising the same document twice is wasted effort.

• The recognised text is scrambled. The layout has confused the column detection. Recognising one column at a time is slower but reliable.
• The voice is hard to listen to for long. Several voices are usually installed. Speed and voice choice make more difference to fatigue than people expect, and both are worth adjusting properly.
• Highlighting and a screen reader clash. As with any literacy toolbar, pick one speech source for the task rather than running both.
• It cannot read a protected PDF. Some files block text extraction. Screenshot recognition is the way round it.

People with dyslexia and low vision, and anyone whose work involves documents that arrive as scans. It is a common Access to Work recommendation where the barrier is the state of the paperwork rather than the person's reading.